University of Southern Queensland
Understanding the impact of privacy concerns and trust on social
networking sites: Analysing user intentions towards willingness to
share digital identities
A Dissertation submitted by
Sanjib Tiwari, MIT, B.Tech (IT)
For the award of
Master of Business Research
2012
ii
Certificate of Dissertation
I certify that the ideas, investigations, analysis, results, discussions, and conclusions
reported in this dissertation are entirely my own work, except where otherwise
acknowledged. I also certify that the work is original and has not been previously
submitted for any other award, except where otherwise acknowledged.
------------------------------------------------------------ Date:…../……/… …
Signature of Candidate – Sanjib Tiwari
Endorsement
------------------------------------------------------------- Date:…../……/… …
Signature of Principal Supervisor – Dr Jianming Yong
------------------------------------------------------------- Date:…../……/… …
Signature of Associate Supervisor – Dr Michael Lane
iii
Abstract
Participation in social networking sites (SNS) has dramatically increased in recent
years. SNS focus on building online communities of people who share interests
and/or activities, or who are interested in exploring the interests and activities of
others. This study examines the experiences of SNS users, and explores how the
depth of their experience and knowledge of the Internet, trust and privacy concerns
impact upon their individual willingness to share information about their own
identity with other users on social networking websites. An acceptance model is
proposed that incorporates cognitive, as well as affective, attitudes as primary
influencing factors on user attitudes and behaviour which, in turn, are driven by
underlying beliefs, perceived levels of privacy and trust, attitudinal experiences and
knowledge, as well as a willingness to share.
The proposed conceptual model for this study is derived from the literature review
and Theory of Planned Behaviour. This model explains how people experience
different levels of motivation about sharing knowledge and seeking information from
other members which, in turn, leads to a divergence in both intentions and
behaviours within virtual communities. The model shows excellent measurement
properties and establishes two distinct constructs—specifically, the need for
perceived levels of privacy, and the need for established levels of trust within SNS.
This study is based on quantitative methodology and uses a structural equation model
to test the construction of the model and its hypothesis. The data for this study were
collected from a Facebook forum, with a sample size of 155 SNS users.
The main theoretical contribution of this study is to provide greater understanding
and new insights into privacy concerns and trust, in so far as these factors impact
upon SNS users‘ willingness to readily share information regarding their digital
identities. Secondly, this study will enrich the existing literature regarding the inter-
relationship between the extent of SNS users‘ length and depth of experience as
Internet users, as this impact upon their willingness to share identity-based
information.
iv
List of publications during the MBSR study
period
Tiwari, S, & Yong, J 2011, 'Understanding Impact of Privacy Concerns and Trust in
Social Networking Sites', paper presented to 5th
International Conference on ICT for
Development and Education, Kathmandu, Nepal, November 16-17, 2011.
Yong, J, Tiwari, S, Xiaodi, H & Qun, J 2011, 'Constructing Robust Digital Identity
Infrastructure For Future Networked Society', paper presented to 15th International
Conference on Computer Supported Cooperative Work in Design (CSCWD),
Lausanne, Switzerland, 8-10 June, 2011.
v
Acknowledgements
I sincerely wish to thank the following people for their encouragement and help in
the researching and writing of this thesis.
I would especially wish to thank my supervisors Dr Jianming Yong for his patience,
his professional advice and his faith in me to complete this thesis. Thanks also to my
associate supervisor, Dr Michael Lane, for his suggestion to explore online social
networking sites and his great enthusiasm for his support for this project.
I also thank my good friends Ramesh Tiwari, Arjun KC, Arjun Neupane and Rohini
Prasad Devkota for their unwavering support and encouragement for this
dissertation.
I remember all those who have contributed directly or indirectly to the successful
completion of my study. In particular, special thanks go to all of the survey
respondents for taking the time and effort to participate in the study.
Lastly, I would like to acknowledge my family and my lovely and caring wife Sunita
for their constant care, love and encouragement. I owe much to them for where I
stand today.
Sanjib Tiwari
vi
Table of Contents
CERTIFICATE OF DISSERTATION ............................................................................... II
ABSTRACT .......................................................................................................................... III
LIST OF PUBLICATIONS DURING THE MBSR STUDY PERIOD .......................... IV
ACKNOWLEDGEMENTS ................................................................................................. V
TABLE OF CONTENTS .................................................................................................... VI
LIST OF TABLES ............................................................................................................ VIII
LIST OF FIGURES ............................................................................................................. IX
LIST OF ACRONYMS ....................................................................................................... XI
CHAPTER I: INTRODUCTION ......................................................................................... 1
1.1 BACKGROUND TO THE STUDY ....................................................................................... 1
1.2 FOCUS AND MOTIVATION .............................................................................................. 3
1.3 STATEMENT OF PROBLEM .............................................................................................. 5
1.4 GOAL AND RESEARCH OBJECTIVES ............................................................................... 7
1.5 METHODOLOGY ............................................................................................................. 8
1.6 OUTLINE OF REPORT ...................................................................................................... 8
1.7 CONCLUSION ................................................................................................................ 10
CHAPTER II: LITERATURE REVIEW.......................................................................... 11
2.1 INTRODUCTION ............................................................................................................ 11
2.2 SOCIAL NETWORKING SITES ........................................................................................ 11
2.2.1 Web 2.0 and Social Networking ........................................................................... 11
2.2.2 Background of Social Networking Sites ............................................................... 16
2.2.3 Defining Social Networking Sites ......................................................................... 21
2.2.4 SNS Knowledge and Experience .......................................................................... 22
2.3 INTERNET PRIVACY ...................................................................................................... 24
2.3.1 Privacy Concerns ................................................................................................. 27
2.4 TRUST........................................................................................................................... 30
2.5 DIGITAL IDENTITIES ..................................................................................................... 33
2.5.1 Willingness to Share Digital Identity ................................................................... 36
2.6 GAPS IN THE LITERATURE ............................................................................................ 37
2.7 CONCLUSION ................................................................................................................ 38
CHAPTER III: THEORETICAL SUPPORT AND CONCEPTUAL MODEL ............ 40
3.1 INTRODUCTION ............................................................................................................ 40
3.2 RELATED SOCIAL THEORIES ........................................................................................ 40
3.2.1 Theory of Reasoned Action (TRA) ........................................................................ 42
3.2.2 Technology Acceptance Model (TAM) ................................................................. 43
3.2.3 Theory of Planned Behaviour (TPB) .................................................................... 43
3.3 RESEARCH QUESTION (RQ) ......................................................................................... 45
3.4 CONCEPTUAL MODEL .................................................................................................. 46
3.5 CONCLUSION ................................................................................................................ 50
CHAPTER IV: RESEARCH DESIGN AND METHODOLOGY .................................. 52
4.1 INTRODUCTION ............................................................................................................ 52
vii
4.2 OBJECTIVE ................................................................................................................... 53
4.3 RESEARCH DESIGN ...................................................................................................... 53
4.4 RESEARCH PHILOSOPHY AND PARADIGM .................................................................... 55
4.5 RESEARCH APPROACH ................................................................................................. 57
4.6 QUESTIONNAIRE DESIGN ............................................................................................. 59
4.7 DATA COLLECTION AND SAMPLE SIZE ........................................................................ 64
4.8 ETHICAL CONSIDERATIONS ......................................................................................... 66
4.9 DATA ANALYSIS .......................................................................................................... 67
4.9.1 Descriptive Statistics Analysis ............................................................................. 67
4.9.2 Reliability ............................................................................................................. 67
4.9.3 Validity ................................................................................................................. 69
4.9.4 Factor Analysis .................................................................................................... 71
4.9.5 Structural Equation Modelling (SEM) ................................................................. 71
4.10 CONCLUSION .............................................................................................................. 73
CHAPTER V: RESULTS AND ANALYSIS ..................................................................... 75
5.1 INTRODUCTION ............................................................................................................ 75
5.2 DATA QUALITY AND CHARACTERISTICS OF RESPONDENTS ........................................ 75
5.3 CONFIRMATORY FACTOR ANALYSIS (CFA) ................................................................ 86
5.4 DIMENSIONAL LEVEL ANALYSIS – THE MEASUREMENT MODEL ............................... 92
5.5 DESCRIPTIVE STATISTICS AND CORRELATION FOR ALL VARIABLES .......................... 95
5.6 STRUCTURE EQUATION MODELLING (SEM) ............................................................... 96
5.6.1 Overall Model Fit ............................................................................................... 100
5.6.2 Path Results ........................................................................................................ 101
5.7 CONCLUSION .............................................................................................................. 102
CHAPTER VI: CONCLUSION ....................................................................................... 105
6.1 INTRODUCTION .......................................................................................................... 105
6.2 SUMMARY OF THE STUDY .......................................................................................... 106
6.2.1 Research problem ............................................................................................... 106
6.2.2 Research hypotheses .......................................................................................... 107
6.2.3 Research Methodology ....................................................................................... 108
6.2.4 Conclusions about descriptive demographic data ............................................. 109
6.2.5 Conclusions about SEM model fit ...................................................................... 109
6.2.6 Conclusions concerning Results of Research Hypotheses Tests ........................ 110
6.3 CONTRIBUTION OF STUDY ......................................................................................... 113
6.3.1 Contributions to the Literature .......................................................................... 113
6.3.2 Contributions for SNS Users .............................................................................. 114
6.3.3 Contribution to SNS Developers ........................................................................ 114
6.4 LIMITATIONS OF THE STUDY AND FUTURE RESEARCH OPPORTUNITIES ................... 115
6.5 SUMMARY .................................................................................................................. 117
REFERENCES ................................................................................................................... 119
viii
List of Tables
TABLE 4-1 QUESTIONNAIRE ITEMS AND VARIABLE CODING ................................................. 62
TABLE 5-1 CHARACTERISTICS OF THE RESPONDENTS ........................................................... 86
TABLE 5-2 RELIABILITY STATISTICS ...................................................................................... 89
TABLE 5-3 CONVERGENT VALIDITY OF THE MODEL VARIABLES ........................................... 91
TABLE 5-4 GOODNESS OF FIT INDEXES OF THE MEASUREMENT MODEL .............................. 94
TABLE 5-5 DESCRIPTIVE ANALYSIS AND CORRELATION....................................................... 96
TABLE 5-6 GOODNESS-OF-FIT INDICES OF STRUCTURAL MODEL ........................................ 100
TABLE 5-7 PATH COEFFICIENTS FOR STRUCTURAL MODEL ................................................. 101
TABLE 5-8 RESULTS OF HYPOTHESES TESTING .................................................................... 102
ix
List of Figures
FIGURE 2.1 TIMELINE OF THE MANY MAJOR SNS .................................................................. 20
FIGURE 2.2 DIGITAL IDENTITY: GLOBAL SET OF ATTRIBUTES OF A USER ............................ 33
FIGURE 2.3 THREE TIERS OF DIGITAL IDENTITY ..................................................................... 35
FIGURE 3.1 CONCEPTUAL MODEL - KEY FACTORS OF SNS THAT IMPACT ON WILLINGNESS TO
SHARE DIGITAL IDENTITY .................................................................................. 46
FIGURE 4.1 THE PROCESS OF QUANTITATIVE RESEARCH ....................................................... 54
FIGURE 4.2 THE FLOW CHART OF RESEARCH DESIGN ............................................................ 55
FIGURE 4. 3 THE PURPOSE RESEARCH DESIGN ....................................................................... 55
FIGURE 5.1 Respondent age groups percentiles ..................................................................... 77
FIGURE 5.2 USES OF INTERNET .............................................................................................. 78
FIGURE 5. 3 USES OF SOCIAL NETWORKING SITES ................................................................. 79
FIGURE 5.4 CURRENT NUMBER OF ACCOUNTS IN DIFFERENT SNS ........................................ 80
FIGURE 5.5 NUMBER OF VISITS TO SOCIAL NETWORKING SITES ............................................ 81
FIGURE 5.6 SOURCE OF ORIGINAL KNOWLEDGE ABOUT SNS ................................................ 82
FIGURE 5.7 AVERAGE NUMBERS OF FRIENDS IN EACH SOCIAL NETWORKING ACCOUNT ....... 83
FIGURE 5.8 PURPOSE FOR VISITING SOCIAL NETWORKING SITES ........................................... 84
FIGURE 5.9 EDUCATION STATUS OF RESPONDENTS ............................................................... 85
FIGURE 5.10 MEASUREMENT FIT MODEL ............................................................................... 92
FIGURE 5.11 STANDARDISED STRUCTURE EQUATION MODEL PATH DIAGRAM ..................... 98
FIGURE 5.12 FINAL STRUCTURAL EQUATION MODEL OF IMPACT OF PRIVACY CONCERNS AND
TRUST IN SNS FOR WILLINGNESS TO SHARE DIGITAL IDENTITIES ........................ 99
x
List of Appendices
APPENDIX A ..................................................................................................................... 134
PARTICIPANT INFORMATION SHEET ............................................................ 134
APPENDIX B ..................................................................................................................... 136
ETHICS APPROVAL ............................................................................................ 136
APPENDIX C ..................................................................................................................... 137
INFORMATION AND INFORMED CONSENT STATEMENT (ONLINE
VERSION) .............................................................................................................. 137
APPENDIX D ..................................................................................................................... 138
ONLINE SURVEY ................................................................................................. 138
APPENDIX E ..................................................................................................................... 143
STATISTICAL DATA ANALYSIS DETAILS ..................................................... 143
xi
List of Acronyms
Acronyms Description
C.R. Critical Ratio
CFA Confirmatory Factor Analysis
CFI Comparative Fit Index
CR Construct Reliability
DI Digital Identities
GFI Goodness of Fit Index
IFI Incremental Fit Index
IS Information System
IT Information Technology
IUIPC Internet Users‘ Information Privacy Concerns
NFI Normed Fit Index
PC Privacy Concerns
RMSEA Root Mean Square Error of Approximation
RSS Rich Site Summary
SEM Structural Equation Model
SNS Social Networking Site
SPSS Statistical Package for the Social Sciences
T Trust
TAM Technology Acceptance Model
TLI Tucker-Lewis coefficient
TPB Theory of Planned Behavior
TRA Theory of Reasoned Action
VE Variance Extracted
1
Chapter I: Introduction
1.1 Background to the Study
Since the introduction of Social Networking Sites (SNS), the growth of social
networks online has been both rapid and dramatic, changing the purpose and the
functionality of the Internet. SNS such as Facebook, Myspace, Linkedin and others
are types of online virtual communities that have grown in popularity in recent years
(Hu & Ma 2010). Lo and Riemenschneider (2010) argued that the growing
popularity of social networking sites has created a new stream of inquiry for
academics and practitioners alike, as indicated by the number of papers appearing in
proceedings of conferences and workshops relating to online social networks.
Increasingly, SNS users have integrated these sites into their daily routines for the
purpose of sharing their interests, and communicating with friends and contacts by
posting and exchanging information about themselves (Shin 2010a). For example,
Facebook alone has over 500 million active users (users who have returned to the site
in the last 30 days), with the average user spending more than 55 minutes per day
and each having 130 friends on Facebook (Facebook 2011).
Most SNS have the common goal of connecting users and building new relationships
using the ‗network effect‘. To build these network connections, users must be willing
to provide information by populating their profile with personal information so that
their friends and acquaintances can search for, identify, and reach them, and then
interact (Fuchs 2010). According to Thelwall (2009), a user profile is a list of
identifying information that can include a person‘s real name, or a pseudonym. It
also can include photographs, birthdays, and the location of their hometown, their
2
religion, ethnicity, political views and personal interests.
Valenzuela, Park and Kee (2009) argued that sharing authentic and genuine details
about a user‘s real identity in this way encourages users to establish trust in order to
develop new relationships. For example, Ellison, Steinfield and Lampe (2007b)
found that Facebook users who shared their profiles on SNS had gained more
friendship connections in their network.
As online social networking sites are considered personal spaces, their usage is often
driven by friendship and relationships, in addition to providing an outlet for creative
and personal expression. Many of these sites are classified as private, which suggests
to the users that they exercise control over who has access to the information on their
site, and gives the illusion of privacy. This study explores whether or not this
expectation of privacy over a user‘s personal space influences how online social
network users perceive their willingness to share information.
Lo (2010) argued that privacy concerns pose a problem, since past research has
consistently revealed that online users are generally concerned about the privacy of
their personal information. Moreover, Dwyer, Hiltz and Passerini (2007) have shown
that, in contexts related to information, privacy and trust are two of the most salient
beliefs affecting people‘s intentions to release information. Further, Luo (2002) built
a trust-privacy framework which suggests that knowledge plays an important role in
determining behaviour in situations in which potential privacy concerns are involved.
He also found that users exhibit individual trust regarding SNS themselves and trust
beliefs about other SNS users. Lo and Riemenschneider (2010) found that a model
3
involving trust and privacy concerns would be an appropriate lens through which to
examine the phenomenon of personal information disclosure on SNS.
Dinev and Hart (2006) observed that users with Internet literacy have differing
perspectives on privacy concerns and trust. Previous research studies found that
privacy concerns exist in electronic commerce sites, but only a few studies have
focused on how concerns about privacy and trust influence users‘ willingness to
share aspects of their identities within SNS. Thus the main objective of this study is
to determine:
To what extent do privacy concerns and trust influence users’ willingness to share
digital identities (information) on social networking sites?
1.2 Focus and Motivation
Interest among SNS researchers has been wide-ranging, with prior research focused
on such topics as social capital (Ellison et al. 2007b), privacy concerns (Gross &
Acquisti 2005), trust (Fogel & Nehmad 2009) the differences between users and non-
users (Hargittai 2008), and identity management (DiMicco, Millen, Geyer, Dugan,
Brownholtz & Muller 2008).
Recent research by Lo (2010) found that many users have major concerns about
privacy and trust as they relate to the use of SNS. However, a number of studies
found similar concerns about trust and privacy had effected the willingness of some
individuals to use electronic commerce, particularly regarding online transactions
(Pavlou & Fygenson 2006; Peter 2001; Rosenblum 2007; Xu 2009); but only a few
studies have explored the privacy-trust issue in relation to people‘s willingness to
4
share personal information within SNS (Fogel et al. 2009; Shin 2010a). Luan, Fung,
Nawawi and Hong (2005) found that users with more internet experience and
knowledge were more willing to communicate and exchange information online.
Fogel and Nehmad (2009) pointed out that users with more experience were
positively disposed towards trusting the existing levels of security within SNS
sufficiently to enable their participation. Moreover, Shin (2010a) found that once a
user trusted SNS, they uploaded their personal data—seemingly without any
concern—and shared this information among their friends without limits. DiMicco,
Millen, Geyer, Dugan, Brownholtz and Muller (2008) established that users with less
Internet knowledge were more likely to place the most trust in SNS sites as they
believe that only friends could view the information that they uploaded. Additionally,
Kim, Steinfield and Lai (2008) found that users with less Internet knowledge were
primarily concerned with using SNS when they did access the internet, and were
more likely to share information than were more experienced users. However, they
also found that once users are equipped with better knowledge of security and
privacy issues, they will be better able to assess the trustworthiness of websites.
Similarly, Lo (2010) determined that users with high levels of SNS literacy were
more likely to exhibit concerns about privacy. However, Dwyer, Hiltz and Passerini
(2007) argued that, in practice, users often ignored privacy issues as they related to
SNS. From the above study it was found that users with substantial SNS experience
and knowledge had different attitudes about their willingness to share aspects of their
identity with other SNS participants than did users with only limited Internet and
SNS experience.
5
Finally, Lo (2010) suggested, according to the study of Fogel and Nehamad (2009),
that the majority of SNS users were university students. Students with the highest
levels of computer literacy were more likely to be concerned about privacy, whereas
students with less-developed computer literacy were less concerned about privacy.
Hence, this study suggested that user SNS knowledge does, in fact, influence
concerns about privacy. However, these researchers did not study how the users‘
experience with SNS influenced their willingness to share information with other
users. Hence, this study is important in order to understand the influence of privacy
concerns and trust upon SNS users‘ willingness to share personal information on
SNS sites. In practice, it has been found that users are sharing their digital identities
in SNS, despite media reporting about breaches in SNS security.
This indicates either that users were unaware of the importance of maintaining
privacy in SNS, or that the users were willing to share their digital identities,
regardless of privacy concerns. This particular issue motivates the proposed research.
1.3 Statement of Problem
As the use of the Internet increases, and as most people‘s time schedules become
busier, many software companies have developed SNS in order facilitate the sharing
of information in a virtual world. This helps people to reach and connect with their
friends and family, with whom they might not have the time or opportunity to
connect by other means. Most SNS are based upon open platforms within which
anyone can join and create a profile. However, in order to use SNS, it is not
necessary that everyone has an extensive knowledge of SNS and how they work.
Boyd (2008) found that some users frequently updated their daily routines and future
6
plans in their SNS profile. Fogel and Nehamad (2009) discovered that most
university students were intentionally sharing their daily routine in order to impress
their colleagues. Also, they found that some users were unknowingly or
unintentionally sharing their personal information on SNS sites.
Dwyer, Hiltz and Passerini (2007) found that privacy concerns and trust were major
issues for users of SNS. As the world has become more digitised, the protection of
the privacy of the user has become more complicated. Nysveen and Pedersen (2004)
found that privacy concerns in SNS were similar to those found for users of
electronic commerce in terms of concerns by users about providing personal
information. Pavlou and Fygenson (2006) found that customers who were using both
SNS and online transactions had more privacy concerns regarding online transactions
than they did about SNS. Furthermore, they found that despite their concerns about
security, users were still willing to share their personal information within SNS.
However, they were unable to identify the reasons why users were more willing to
share their personal information within SNS.
Previous studies conducted by Lo (2010) concerning privacy concerns in SNS
suggest that knowledge and experience may affect the perception of users regarding
privacy issues as they relate to the sharing of personal information. But his study did
not justify whether users‘ SNS knowledge and experience play critical role to
influence users‘ privacy concerns and trust to share personal information in SNS.
Dwyer, Hiltz and Passerini (2007) found that trust was another factor that determined
users‘ willingness to share personal information in SNS. Debatin, Lovejoy, Horn and
7
Hughes (2009) contend that the issue of trust in SNS could be separated into two
distinct aspects: trust about SNS itself; and trust of other SNS users. According to
Boyd (2011), users with Internet experience might easily trust SNS because they
have knowledge of security and privacy settings within the Internet which, in turn,
make users feel more secure when using SNS. Furthermore, Binder, Howes and
Sutcliffe (2009) also found that frequent users of SNS tend to build trust in the SNS
itself, rather than with their friends. However, Lo (2010) found that users had more
trust in their SNS friends than SNS themselves because of levels of knowledge and
belief in their network of friends.
However, few studies have outlined the important of users‘ trust on using new
systems, and none of the above studies investigated the impact of trust and a
willingness to share their personal identity in SNS. Thus, this study focuses on user
behaviours and attitudes towards privacy concerns and trust issues, as these factors
influence user willingness to share information about users‘ own digital identities.
1.4 Goal and Research Objectives
Based on the previously-outlined research problem, the main research question of
this study is:
To what extent do privacy concerns and trust influence users’ willingness to
share information about their digital identity within Social Networking Sites?
Therefore, the main research objectives that underpin the general research question
of this study are:
To examine the impact of users' experience with SNS upon their willingness to
8
share their digital identities.
To examine the influences of privacy concerns as they relate to the trust needed
for users to share their digital identities within SNS.
To examine the influence of trust of SNS upon users‘ willingness to share their
digital identities.
To examine the impact of privacy concerns upon the willingness of users of SNS
to share their digital identities.
1.5 Methodology
Chapter Four provides details of the methodology used in this study and provides
justification for its use and implementation. An online questionnaire was used to
collect data and a survey link was posted in the SNS forum. The instrument was
adopted from previous study, so validity and reliability for this study purpose are
assured. The exploratory and confirmatory factor analyses using SPSS 19.0,
Structural Equation Modeling using AMOS 19.0 was performed to examine the
goodness-of-fit indices of the various measurement and structural models.
1.6 Outline of Report
This thesis consists of six chapters. The first chapter (Introduction) provides
background to the research, and briefly outlines the gaps in the research that this
study seeks to explore. It introduces the research question and the research objectives
and outlines the methodology chosen to explore the research question.
Chapter Two (Literature review) reviews the literature in the areas of SNS, digital
identity, privacy concerns and trust. This study also explores the background, history
9
and development of SNS. Finally, this study investigates the demographic data that
users are willing to share in SNS.
Chapter Three (Theoretical support and conceptual model) lays out the research
framework for this study. This chapter reviews the relative theories used in
information systems that relate to this study. Emerging from the literature review and
theoretical support, a research conceptual model is presented and four hypotheses
identified for further research.
Chapter Four (Research design and methodology) provides an outline and
justification behind the methodology and collection of data for this study. It presents
the methods, a description of the sample and an outline of the research process.
Chapter Five (Results and analysis) summarises and presents the findings of the
study, using four hypotheses as the organising structure for analysis, and the key
findings are identified and summarised.
Chapter Six (Findings, recommendations and conclusion) discusses the key findings
from the data and provides conclusions to the study. The motivations behind social
networking users are discussed and outlined, the perceptions of SNS are detailed, and
issues such as users and their perceptions relating to trust and privacy in SNS are
outlined. Users‘ willingness to share information regarding their identity within SNS
is discussed and a new model is presented. The findings of this study reveal that the
willingness by users of SNS to share their digital identities is determined by their
individual perceptions about of privacy and trust within SNS. This chapter also
10
discusses the implications of theory and practice that arise from the results of this
study. The limitations of this study are also addressed, and recommendations are
made for areas of future research.
1.7 Conclusion
This chapter outlines the background to the research question by introducing social
networking sites and describing their rapid growth in the Internet market place.
Owing to the relative newness of social networking, there is little published research
on the topic of privacy concerns and trust as they relate to user intentions and
actions. Hence, this research breaks important new ground by exploring exactly how
users‘ willingness to share digital identities is impacted upon by their own issues of
privacy and trust as they relate to the use of SNS. The research problem that guides
this study and its research is presented, and an overview of the study methodology is
described. The outline of the report is detailed and the limitations of the study
acknowledged. A review of the literature begins in Chapter Two.
11
Chapter II: Literature Review
2.1 Introduction
This section contains a review of the existing literature that is related to this project,
beginning with a description of the concept of Web 2.0, and provides a background
about Social Networking Sites beginning with the very first Social Networking Site,
sixDegrees (www.sixdegree.com), up to the creation of the most popular current
SNS, namely, Facebook. The focus of the literature review was to understand SNS,
its usage, and to explain the public‘s increased participation in SNS and the impact of
privacy and trust upon users‘ participation in social online networks. The literature
review also shows how users can create their public profiles and share their digital
identities with other users.
Gradually-increasing numbers of Internet users and social networking sites has
created the issue of privacy concerns and trust. So the main objective of this research
is to review users‘ privacy concerns and trust that effect users‘ willingness to share
digital identities on SNS.
2.2 Social Networking Sites
2.2.1 Web 2.0 and Social Networking
The creation of Web 2.0 has facilitated an entirely new type of communication that
became increasingly important to society. Web 2.0 is the popular term for advanced
internet and applications and includes blogs, Wiki, RSS, and Social Networking Sites
(Lai & Turban 2008). According to Anderson (2007) and Madden and Fox (2006),
the concept of Web 2.0 was created by Dale Dougherty and O‘Reilly Media Inc in
2004, and the term has become more popular and its use continued following ‗dot
12
com‘ in recent years. Sir Tim Berner-Lee, the creator of the World Wide Web,
criticised the term Web 2.0 as he believed it was nothing more than a
fully-implemented version of the original Web. To support his view, he reportedly
argued that Web 2.0 was based on Web 1.0 standards, and the purpose of both Web
1.0 and Web 2.0 was to create connections between people (Paul et al. 2007).
However, Trembath (2011) argues that what differentiates Web 2.0 from more
traditional IT, including the Web, is not just one attribute but, rather, a set of
characteristics that together give shape to this new class of technologies, and at the
same time provides the field of IT research and practise with some interesting
challenges.
In practice, it is generally accepted that many standards that underpin Web 2.0 have
been derived from the traditional Web, and that Web 2.0 has a much more social
orientation than Web 1.0. Lai and Turban (2008), for instance, argued that the
combination of user-generated content, its collaborative nature and the significant
emphasis on Social Networking Sites make Web 2.0 more advanced than the
traditional Web. Rather than being defined with reference to a list of specific
applications and services, Web 2.0 is usually described as embodying a set of
principles and practices. The associated applications and services usually have
defining characteristics that enable users to create online content, access collective
intelligence and access network-enabled interactive services (Madden et al. 2006). At
the core of Web 2.0 is a sense of participation and ‗collaborations, contributions and
communities‘ (Paul et al. 2007, p. 14) and there are a range of websites (e.g.
Wikipedia, YouTube and Blogger) that support Web 2.0 activities such as
collaboration, media sharing and blogging. The focus of this study, however, is on
13
social networks.
The most popular types of Web 2.0 applications that have developed in recent years
are online Social Networking Sites (SNS) or virtual communities, in which
membership continues to grow exponentially (Lai et al. 2008). SNS such as
Facebook, MySpace, Twitter, LinkedIn, Google Plus, Hi5, and Friendster are new
forms of self-representation and communication, and imply a social behaviour that is
different to the real world (Bonhard & Sasse 2006). Since their introduction, these
SNS have not only attracted millions of users, but have become an essential part of
the users‘ everyday activities—a parallel universe that, in the virtual world, satisfies
the human need for sociability (Ganley & Lampe 2009). Social networking sites
generate billions of dollars in revenue and are being increasingly used in marketing
and advertising campaigns. However, very little research has been carried out to
investigate the factors that influence the usage of SNS, as suggested by
Gangadharbatla (2010).
The Pew Internet Project defined online social networks as ‗spaces on the Internet
where users can create a profile and connect that profile to others (individuals or
entities) to create a personal network‘ (Lenhart 2009, p. 1). Social networking sites
enable individuals to connect with their friends and colleagues, as well as to form
new associations through participation in online groups. Yahoo! Groups (Yahoo
2011), for instance, is promoted as a place ‗where people get to know each other and
stay informed‘. Alternatively, Facebook (2012) declares that it is their mission to
‗give people the power to share and make the world more open and connected‘.
Group members typically have online access to group features such as forum
14
postings, photo albums, shared links and group birthday calendars, as well as
individual features such as personal profiles. Plant (2004) argued that memberships
of online communities satisfied two basic human needs: the need to be connected to
others; and the need to acquire knowledge. Supporting this argument, a
recently-completed study by Fox and Purcell (2010) found that individuals seeking
health information benefited significantly from tapping into the pool of user-
generated information and the emotional support provided within online social
networks (Fox & Purcell 2010).
The amount and scope of information that SNS users freely reveal is stunning and
constitutes a highly attractive database and profiling source for different interest
groups, ranging from marketers to recruiters, private detectives, public authorities,
and hackers. Information technology experts characterise Web 2.0 social networks as
‗attractive targets for those with malicious intent‘, because each site offers a huge
user base, sharing a common infrastructure, and the information that users willingly
supply is highly valuable (Mansfield-Devine 2008). The average user‘s profile
contains information about their home address, their pet‘s name, where they went to
school, their mother‘s maiden name, their likes and/or dislikes, interests, hobbies and
other family details just the kind of information used for security or ‗lost password‘
questions that are required routinely, for example, by online banking services. Many
participants also provided detailed information about their interests, sometimes
including their political and sexual orientations or intimate portraits of their social or
inner lives (Gross et al. 2005). Every now and then, problems related to privacy, trust
or security issues on social network sites are reported in the media. For instance, in
May 2008 the social networking website Bebo admitted that a ‗bug‘ in its systems
15
enabled users to view other people‘s private information. Phone numbers and
addresses were made available as some of Bebo‘s 40 million users found themselves
randomly switched to other people‘s accounts (Eriksen 2008). Evidence from many
other online social networks indicate that despite these reported breaches of security
in the past, nevertheless, millions of social network users do not hesitate to share
their thoughts, experiences, images, files, videos, and links in an environment that is
largely devoid of security standards and practices.
It has also been argued that social networking sites are well-suited to meet
information and connection needs because they foster the development of sparse,
unbounded networks that encourage the formation of weak ties (Wellman 1997).
‗Sparseness‘ refers to the number of contacts that members have with each other,
while ‗boundedness‘ refers to the percentage of members‘ ties that stay within the
boundaries of the network. ‗Tie strength‘ describes frequency of contact, social
closeness, reciprocity and the degree of voluntary involvement (Granovetter 1973;
Wellman 1997). Weak ties typically help individuals reach out to various
information sources and resources, and are more likely to exist between
acquaintances. Within this context, the social aspects of online networking are
becoming increasingly important (Paul et al. 2007). Strong ties, on the other hand,
typically provide companionship and emotional support, and are most likely to exist
between family and close friends. They are founded on considerable trust and
support. While it is generally accepted that it is the more traditional communities,
such as those described by Tonnies (1887) and Durkheim (1893), that typically foster
the development of strong ties, a recent study conducted by the Pew Research Centre
in 2009 found that a high percentage of people surveyed (85%) used online social
16
networking applications to interact with people whom they already knew offline
(Lenhart 2009). On that basis, it could be argued that strong ties, as well as weak ties,
can be maintained online. Lai and Turban (2008) argued that online social
networking sites have experienced a significant increase in popularity since the
emergence of Web 2.0 applications and services. Moreover, the Pew Research
Centre reported that the number of American adult internet users who had reported
using a social networking site had quadrupled within the preceding four years, from
2005 (8%) to 2008 (35%), and had almost doubled from 2008 to 2011 (65%)
(Lenhart 2009).
Disclosing personal information on the Internet presupposes trust, because the user
does not know whether their personal information may be used in ways that the user
is not able to foresee, and as a result may cause potential harm to the user, or lead to
unwanted future solicitations or hijacking of one‘s online identity for personal use
(Milne & Culnan 2004). Obviously, social networking takes place within a (largely
unwarranted) context of trust. Consequently, the question arises why social network
users are being so trusting, despite their concerns about privacy. Scant research has
considered the interrelationships between privacy concerns, trust, social networks,
and the Web 2.0 environment. In this study, analysis is conducted on the role that
privacy concerns and issues of trust play vis-a-vis SNS from an Internet network
governance perspective that integrates concepts of behavior intention theories such
as the Theory of Planned Behavior.
2.2.2 Background of Social Networking Sites
Online social networks are a fast-growing phenomenon and are emerging as the
Web‘s top application (Chiu, Cheung & Lee 2008). SNS have become a computer-
17
mediated communication medium in the Internet world (Ahn, Han, Kwak, Moon &
Jeong 2007). At the most basic level, SNS are online communities based on a social-
circle network model, in which people build their own profile and create a network
of connections with other participants. Boyd and Ellison (2008, p. 211) defined SNS
as web-based services that allow individuals to construct a public or semi-public
profile within a bounded system, to articulate a list of other users with whom they
share a connection, and then to view and traverse their list of connections and those
made by others within the system. The visibility of a user‘s profile on different SNS
determines to what extent SNS users can view other users‘ profiles and, normally, a
SNS like Facebook has default settings, although these can be configured by
individual users (Hoy & Milne 2010; Staksrud & Lobe 2010).
At the time of writing, there are hundreds of SNS offering different types of services
to individuals and groups with shared interests. These SNS display great diversity in
user bases across genders, age groups and specific geographical regions. The first
wide-scale online social network ‗sixdegrees.com‘ was created by Andrew
Weinreich in 1997 (Albrechtslund 2008; Boyd et al. 2008; Livingstone 2008) and
was closed in 2000 due to decreased popularity, although the site had features of
profiles, friend circles and messaging (Boyd et al. 2008; Harrison & Thomas 2009;
Shafie, Mansor, Osman, Nayan & Maesin 2011). It worked in such a way that a
person who registered at the site could list up to ten friends only. Those friends were
supposed to join and list ten friends each of their own, and so on. The site was used
for apartment searches, job hunts, searches for medical specialists or lawyers, and
even finding old high school colleagues (Boyd et al. 2008; Buote, Wood & Pratt
2009).
18
Boyd and Ellison (2008) present the history of SNS since 1997 when
‗SixDegrees.com‘ was launched. They document the timeline up to late 2006,
detailing the release dates of some of the major SNS, including online communities
such as LunarStorm, AsianAvenue and QQ that have been re-invented and re-
launched with SNS features (see Figure 2.1). Among the SNS found in the list,
Facebook, MySpace, LinkedIn, Hi5, and Friendster are the most notable and well-
known sites (Boyd et al. 2008). Also, taking a closer look at these SNS, it is not
surprising that many of them resemble each other in their design and layout, as well
as similarities in the features offered to users.
In 1999, LiveJournal integrated extra features such as guest books and diary pages
(Harrison et al. 2009; Shafie et al. 2011); and in 2003, Linkedin was officially
launched (LinkedIn 2011). By way of comparison, a relatively small Social
Networking Site, My Connected Community (mc² 2011), had 17,049 registered
members in 2002 and 81,851 registered members in 2010. However, Social Network
Sites began to be popular with the general public with Friendster, MySpace and
Facebook (Shafie et al. 2011). Alternatively, Facebook and MySpace are the two
largest SNS in use today, as measured by the total number of registered users in 2010
and 2011 (Boyd 2011; Shin 2010b). In 2004, The Times newspaper predicted that
Facebook would reach 55 million citizens in the near future. Now, Facebook has
over 500 million active users, and approximately half of them log on to their
Facebook account every day (Facebook 2011). On the other side, MySpace has over
100 million monthly active users worldwide and more than 68 million totally unique
users in the USA (MySpace 2011). Other popular social networking sites in 2011
include Twitter, Flixster, Linkedin, and Orkut (Boyd 2011). According to Wu,
19
DiMicco and Millen (2010), there are a number of SNS created by individual
organisations only for the use of their staff. Even though the number of members
may vary, most Social Networking Sites have comparable features. Yahoo! (Yahoo
2011) and My Connected Community (mc² 2011), for instance, provide access to
forums, photo albums, group event calendars, shared links and member polls. They
also allow members to customise their profile pages.
20
Figure 2. 1 Timeline of the many major SNS
(Source: Boyd and Ellison (2008))
21
2.2.3 Defining Social Networking Sites
Broadly, a social network can be defined as a set of actors and a set of ties
representing some relationship, or lack of relationship, between the actors (Brass,
Butterfield & Skaggs 1998). Actors in a social network (people, organisations, or
other social entities) are connected by a set of relationships, such as friendship,
affiliation, financial exchanges, trading relations, or information exchange. SNS uses
computer support as the basis of communication between its members (Andrews,
Preece & Turoff 2001; Andrews 2002). SNS are organised around users, and provide
a basis for maintaining social relationships, for finding users with similar interests,
and for locating content and knowledge that has been contributed or endorsed by
other users (Mislove, Marcon, Gummadi, Druschel & Bhattacharjee 2007).
Web-based social networks provide different means for users to communicate, such
as e-mail, instant messaging services, blogging, and photo/video-sharing. Since
1999, hundreds of online social networks have been launched, with similar
technological features that support a wide range of interests and practices (Ellison &
Boyd 2007a).
Social network sites provide different sets of services, and can be oriented around
work or business related contexts (e.g., XING), romantic relationship initiations (the
original goal of Friendster), or they could aim at connecting those with shared
interests such as music (e.g., MySpace) or the college student population (e.g.,
StudiVZ, or the original launch of Facebook). On the other hand, LinkedIn provides
a service that helps users exchange information and opportunities with broader
networks of professionals. Most online social networks support the maintenance of
22
already-existing social ties, but there are also networking services that support the
formation of new connections with strangers, based on shared interests, political
views, or activities. Some online social networks are directed at diverse audiences,
whereas others attract people based on common interests or shared racial, sexual,
religious, or nationality-based identities (Boyd et al. 2008).
Drawing on Boyd and Ellison (2008) for the purposes of this study, online social
networks are defined as web-based services that allow individuals to (1) create a
public or semi-public profile for themselves within a bounded system, (2) indicate a
list of other users with whom they are connected, and (3) view and traverse their list
of connections and those made by other users within the system. The types and
specific names of these connections may vary from network to network. However,
for this study, SNS means any Social Networking Site that allows any public user to
create their profile and share information. With the exponential growth of social
networking, one may expect that users‘ experience on SNS is increasing.
2.2.4 SNS Knowledge and Experience
DiMicco, Millen, Geyer, Dugan, Brownholtz and Muller (2008) advocated that SNS
Knowledge is a part of Internet knowledge or experience. Novaka, Hoffman and
Yung (2000) argued that Internet experience is usually defined as general experience
with web sites, and not as experience with one particular web site. Additionally,
Chang and Chen (2008) found that by visiting several web sites and using various
value-added services, users will increase their knowledge in general Internet
experience. The length of internet experiences or knowledge for a user may play a
critical role in their evaluation of their SNS experience, as suggested by Binder,
23
Howes and Sutcliffe (2009). Moreover, Gefen, Karahanna and Straub (2003) and
Dahlen (2002) found that through repeated usage of a product or performance of a
task, people became more experienced and gained knowledge. Therefore, for the
user, the more experience of using the Internet they have, the greater is their ability
to build more knowledge about the Internet.
Dinev and Hart (2006) suggested that an internet experience is similar to a SNS
experience, which is measured in length of SNS usage and frequency of visits. A
user‘s degree of Internet experience has some bearing upon how quickly the user
might learn to navigate in an unfamiliar information space. Years of online
experience have proven to be a significant predictor for users in relation to their
adoption and effective use of online commerce in all forms and sharing information
(Flicker 2004). Therefore, the more familiar a user is with the SNS website, the more
likely it is that they will use the different applications available on a SNS web site,
seemingly without any concerns.
A prior experience has been found to be an important determinant of behaviour
(Ajzen & Fishbein 1980). According to Taylor and Todd (1995a) there are some
significant differences in the relative influences of the determinants of usage
depending on experience. However, there was a stronger link between behavioural
intention and behaviour for the experienced users explained by Ajzen and Fishbein
(1980). Moreover Ajzen and Fishbein (1980) found that experienced users employed
the knowledge gained from their prior experience in order to inform their intentions
(Ajzen et al. 1980). In addition, Eagly and Chaiken (1993) suggested that knowledge
gained from past experiences would help to shape users‘ intention, in part because
24
experience makes knowledge more accessible in memory.
Novaka, Hoffman and Yung (2000) identified that experience with the Internet is
among the most important factors that predict online shopping behaviour.
Furthermore, Luan, Fung, Nawawi and Hong (2005) found that users with more
experience and knowledge in using the Internet were more willing to experiment
with online shopping. Their empirical findings reveal that regular visitors to the
Internet were more knowledgeable and experienced and likely to perceive an absence
of certainty in online relationships. They also found that this increased level of trust
is positively related to the user‘s Internet knowledge and experience. Hence, this may
imply that users with high levels of experience and knowledge in using the Internet
and SNS are more willing to use SNS to share information about their digital
identities.
2.3 Internet Privacy
Bandeis and Warren (1890) defined privacy in general terms as the right to be left
alone. This definition has been the basis of the privacy debate that has taken place in
industrialised nations since the beginning of the information era. Privacy in terms of
the Internet is defined as personal information that an individual deems important
and unattainable either by the general population or government surveillance or by
any intrusion (Hodge 2006; Richards 2006; Timm & Duven 2008).
Lo (2010) argued that privacy has been considered to be the greatest issue facing
Internet users today. A major reason people cite for not using the Internet is a fear
about privacy and security (Metzger & Docter 2003; Paine, Reips, Stieger, Joinson &
Buchanan 2007; Ramgovind, Eloff & Smith 2010). According to Lach‘s (1999)
25
survey, most of the online they studied thought that there should be laws to protect
online privacy in Internet.
According to Hodge (2006), when contemplating issues of privacy there are two
important things to keep in mind: the intent of the information shared and the
expectation that it will remain private. A person who willingly posts or shares
information on SNS for others to view cannot assume it is private, because the intent
is to share that particular information (Meredith 2006). On the other hand, Hodge
(2006) argued that when a user adjusts their privacy settings to prevent most users
from viewing his or her information, the user has an expectation that this information
will remain private.
Most Internet sites provide privacy statements, or terms and conditions that apply to
users of the sites: for example both Facebook and MySpace provide a clear privacy
statement and terms and conditions to inform users about the limits of protection that
the site maintains for the information shared, as well as how the site will use the
personal information provided. These privacy policies do not delineate who can
access the information posted on the site, but outline the actions that are taken by the
site‘s administrators. The focus of these privacy statements is to outline what
information will be shared with a third party, but does not address the issue of who
else might access the information that is posted therein. Little is known about
whether individual users read or are aware of privacy settings. However, when
Facebook created its news feed feature, users were outraged that ‗friends‘ would be
informed of their actions on the site. Facebook states that it will do everything
possible to protect the information posted on the site but its creators ‗cannot and do
26
not guarantee that User Content you post on the Site will not be viewed by
unauthorised persons‘ (Facebook 2011). In addition to privacy policies that outline
how Web sites will protect personal information provided to the company, the sites
also outline who is responsible for the information posted in a profile. Facebook
(2011) states, ‗You may not want everyone in the world to have the information you
share on Facebook; that is why we give you control of your information‘. Both
MySpace and Facebook provide advice to parents and users about how to keep the
information shared in the profile protected. MySpace cautions users: ‗Don‘t forget
that your profile and MySpace forums are public spaces‘ (MySpace 2011).
According to Timm and Duven (2008), Chris Hughes, a co-founder of Facebook,
stated that Facebook has provided ways for users to continue to connect online and
that it is up to the user to protect his or her own information by using the tools
provided on the site. The tools provided to social networking site users include a set
of privacy controls that users can adjust to prevent others from viewing all
information shared in a profile. On most sites, the default or automatic settings allow
the profile to be seen by the maximum number of people. On Facebook, the default
setting for a profile is set so that all members of the person‘s network can view the
entire profile. On MySpace, the default setting for a profile is that all users on
MySpace can view a user‘s profile (Timm et al. 2008). The privacy options that are
available for users for other sites vary. On most sites, a user can restrict who can see
their profile, and is given options to create a limited profile that makes parts of his or
her information unavailable to all friends. Although these options are available, many
users do not use these privacy settings (Barnes 2006).
27
2.3.1 Privacy Concerns
Dinev and Hart (2006) argued that an individual‘s Internet privacy concerns reflects
his or her uneasiness about the potential opportunistic behaviour related to his or her
personal information submitted over the Internet. Moreover, Dwyer, Hiltz and
Widmeyer (2008) argued that privacy within social networking sites is often not
expected or remains undefined. Additionally, Nooteboom (2007) and Son and Kim
(2008) suggest that opportunistic behaviours may include a range of actions taken by
others who use or misuse an individual‘s personal information, for example, identity
theft, social engineering to extract one‘s financial information, and spam. Hence,
users with SNS knowledge sometimes worry about the risks associated with sharing
identity information on SNS.
Culnan and Bies (2003) and Xu (2009) argue that the privacy of information varies
according to numerous factors, including industry sectors, culture and regulatory
laws. Privacy concerns about information refer to an individual‘s subjective views of
fairness within the context of information privacy (Davison, Clarke, Jeff, Langford &
Kuo 2003; Malhotra, Kim & Agarwal 2004; Zarsky 2004).
Malhotra, Kim and Agarwal (2004) developed an Internet Users‘ Information
Privacy Concerns (IUIPC) model to examine Internet privacy concerns. They
conceptualised IUIPC as the degree to which an Internet user is concerned about
collection by online marketers (vendors) of personal information; the users‘ control
over the collected information; and the users‘ awareness of how the collected
information is used.
28
With the proliferation of the Internet, Lo and Riemenschneider (2010) believe an
important factor that contributes to shaping the individual‘s Internet privacy concerns
is his or her level of Internet literacy. Further, Luan et al. (2005) argue that users with
high levels of knowledge and experience with the Internet are more willing to
become online shoppers and to place their trust in the sites. However, Hoffman,
Novak and Peralta (1999) and George (2002) found that privacy concerns lead to a
decreased likelihood of online purchases, and that the belief in the privacy of
personal information was associated with negative attitudes towards internet
purchasing.
Lo and Riemenschneider (2010) argued that the precise dynamics of how an
individual‘s extent of Internet literacy influences his or her Internet privacy concerns
was not straightforward. On the one hand, Bateman, Pike and Butler (2011) found
that lower degrees of Internet literacy may elevate concerns about privacy, because
although the individual may be vaguely aware of potential dangers, he or she may
not know how to manage them. On the other hand, higher degrees of Internet literacy
(Sheehan 2002) and increases in education levels (O‘Neil 2001) can likewise raise
concerns about privacy, because the individual is more aware of the dangers.
However, Zhang and Tatipamula (2011) suggested that a more literate individual
may have the knowledge to attempt to minimise the dangers by installing and
updating operating system and browser security patches and fixes, anti-spyware
software and alerts, and other prevention utilities.
Due to this perception of control (i.e. the ability to try to minimise potential dangers),
individuals with higher levels of Internet literacy were expected to be less concerned
29
about their Internet privacy than individuals with lower Internet literacy. However, in
a study of the behavioural intention to conduct online e-commerce transactions,
Dinev and Hart (2006) found that Internet literacy was negatively related to Internet
privacy concerns.
Khosrow-Pour (2007) suggested that willingness to provide personal information
varied depending on the level of privacy offered by policy statements, and that
respondents were most willing to provide information when given strong privacy
statements by a SNS. Moreover, the Theory of Planned Behaviour, which is derived
from the Theory of Reason Action, suggests that beliefs related to behaviour are
prevailing determinants of a person‘s behavioural intentions (Ajzen 1991). As such,
in the context of SNS, a person‘s beliefs about Internet privacy should also be a
determinant of his or her intention to share personal information on a SNS.
However, in general, most consumers who participated in the e-commerce context
were concerned about their online privacy. Findings from a recent study by the USC
Center for the Digital Future indicate that 93% of online shoppers were concerned
about the privacy of their personal information (USC 2011). Meinert, Peterson,
Criswell and Crossland (2006a) suggested that this concern was among the chief
factors discouraging users from shopping online. In line with this reasoning, Dinev
and Hart (2006) found a negative relationship between Internet privacy concern and
the user‘s general willingness to provide personal information in order to conduct
transactions on the Internet. In the context of SNS, the relationship between Internet
privacy concerns and users‘ willingness to submit personal information should be
similar to those extant in the e-commerce context. That is, higher privacy concerns
30
should lead to a reduced willingness to submit personal information to a SNS, as well
as a reluctance to share this information with other SNS users (i.e. friends and
everyone). Hence, higher levels of privacy concerns will result in SNS users having
less positively disposed intentions to share digital identities and trust.
2.4 Trust
Trust has become more important in a high tech environment (Fukuyama 1996). In
the absence of trust, Web sites will most likely exist in an environment devoid of
loyalty. Trust is also important for successful online interactions (Corritore, Kracher
& Wiedenbeck 2003; Liu, Marchewka, Lu & Yu 2005). Mayer, Schoorman and
Davis (2007, p. 712) define trust to mean that one believes in, and is willing to
depend on, another party. Further, Liu, Marchewka, Lu and Yu (2005) define trust as
a perceptual belief or level of confidence that someone respects the intentions, action
and integrity of another party during an online transaction.
Lo and Riemenschneider (2010) suggested that trust for SNS can be compared with
the observed purchasing behaviour of internet users. Hoffman, Novak and Peralta
(1999) implied that the primary reason many people have yet to shop online or to
provide personal information to a vendor is due to a fundamental lack of trust with
online transactions that require the customer to provide credit card and personal
information. However, Pavlou (2003) argued that trust creates positive feelings for
users who partake of online transactions with web retailers, providing expectations
for a satisfactory transaction and thus resulting in a positive attitude towards the
transaction. Therefore, these studies suggest that users with Internet experience and
trust are more willing to regard the sharing of their digital identity on SNS in a
31
positive light.
Liu et al. (2005) proposed a theoretical model that explains how trust influences
consumer behavioural intentions vis-à-vis online transactions. They found strong
support for the positive relationship between the levels or degrees of trust an
individual has with an online business, and the individual‘s own behavioural
intentions. Additionally, Gefen, Karahanna and Straub (2003) suggest that the higher
the customers‘ trust in the web, the less effort customers will exert to scrutinise the
details of the site to assess the authenticity of its services. Research in the field of
electronic commerce has found that trust is strongly related to information disclosure
and was reported as a significant precursor to the disclosure of information online
(Hoffman et al. 1999; Metzger 2004).
Metzger (2004) found that sharing information on SNS with trust was a precondition
for self-disclosure, because it reduced perceived risks involved in revealing personal
identities. Malhotra, Kim and Agarwal (2004), in their IUIPC model, show that
trusting beliefs have a positive effect on users‘ intention to willingly share personal
information on SNS.
Mayer, Davis and Schoorman (1995, p. 474) define trust to mean a state in which
‗one believes in, and is willing to depend on, another party‘. According to McKnight,
Cummings and Chervany (1998) two components of trust are beliefs (i.e. trust
beliefs) and willingness (i.e. trust intentions). Trust beliefs, or ‗factors of perceived
trustworthiness‘ as Mayer et al. (1995) referred to them, are antecedents to trust
intentions according to McKnight et al. (1998) and are comprised of three factors:
32
competence, benevolence, and integrity (Mayer et al. 1995; McKnight, Choudhury &
Kacmar 2003). Specifically, competence refers to the ability of the trustee to perform
the behaviours in the relevant domain expected by the truster. Benevolence refers to
the extent to which the truster believes the trustee cares about and is motivated to act
on behalf of the truster, rather than for the self-gain of the trustee. Finally, integrity
refers to the extent to which the truster believes the trustee is honest and consistently
keeps their promises.
Ajzen and Fishbein (1977), after a review of previous empirical studies about
attitude-behaviour relations, concluded that stronger relationships exist between
attitudes and behaviours when the attitudinal and behavioural entities have high
correspondence. To support this statement, McKnight el al.(1998) explained that a
person may have favourable attitudes toward football games, but unfavourable
attitudes toward attending football games. Thus, measuring general attitudes toward
football games should produce a weak relationship to attendance (the behaviour);
however, measuring the person‘s attitude toward attending football games should
generate a strong relationship with actual attendance. Moreover, Mayer el al. (2007)
suggested that beliefs about an entity should also relate more highly with intention to
behave toward that entity (in this study, with willingness to share identity
information) when the belief and intention entities measured have higher
correspondence.
Finally, Metzer (2004) and Liu et al. (2005) studied privacy, trust and disclosure in
exploring barriers to electronic commerce and found that trust has a positive effect
on users‘ intentions towards willingness to share information; however, these studies
33
did not mention how trust can influence user knowledge and their disposition
towards sharing identity. Therefore, this study concludes that trusting beliefs have a
positive effect on users‘ willingness to share personal information in SNS.
2.5 Digital Identities
Figure 2. 2 Digital Identity: Global Set of Attributes of a User
(Source: Ahn, Ko and Shehab (2008))
There are various definitions of digital identity. The concept of digital identity varies
depending upon its context of use. The term ‗Digital Identity‘ used here describes a
person‘s presence within the Internet as present in various applications, or, in other
words, digital identity is a global set of attributes that make up an online
representation of who and what a user is (Williams, Fleming, Lundqvist & Parslow
2010). It can include access credentials, personal attributes and references. An
individual‘s Digital Identity, as perceived by other people, is made up of material
that the individual manually posts (for example, photographs on Flickr and on one‘s
own web page), but it also is made up of material other people publish to the Internet
about individuals, such as blog posts that mention an individual, and photographs in
which individuals are tagged (Williams et al. 2010). Over the Internet, a user has
numerous access credentials that are issued from different sites; and different or
duplicated personal attributes and references on each site. Identity is an elusive
34
concept: there is no single clear definition. The collective mass of all of the personal
attributes when considered in totality can be judged to mean the users‘ digital
identity as shown in Figure 2.2. In each site where users need to employ some
authentication protocols in order to be active, a user can be represented by a subset of
these attributes. For example, on an auction site such as eBay, a subset of a user‘s
attributes such as username, password, purchase history and credit details represent
the user‘s identity on this site, while for a university or college site, a user‘s identity
may consist of their student ID number, password, class record and Grade Point
Average or GPA (Ahn, Ko & Shehab 2008).
Identity is an important part of the formation of a self-concept. Self-concept is the
totality of a person‘s thoughts and feelings in reference to oneself as an object (Gecas
1986; Zhao 2005; Zhao, Grasmuck & Martin 2008), and identity is that part of the
self ‗by which we are known to others‘ (Altheide 2000, p. 2). The construction of an
identity is therefore a public process that involves both the ‗identity announcement‘
made by the individual claiming an identity, and the ‗identity placement‘ made by
others who endorse the claimed identity, and an identity is established when there is
a ‗coincidence of placements and announcements‘ (Stone 2005). This definition
indicates that the identity of a person is comprised of a set of attributes or properties,
perceived both internally and verified externally, that together make the person
unique (Ellison et al. 2007a).
Digital identity follows this same notion, however, within the context of the Social
Web, the identity of users is bespoke and can be altered by the individual users
(Kamel Boulos & Wheeler 2007; Rowe 2010). Such alterations are possible through
35
functionalities and feature sets on Social Web platforms such as profile pages
(Robards 2010). On such pages users are able to create an identity profile consisting
of their biographical information, which is an Identity for that user (Rowe 2010).
Therefore, in this study, digital identity refers to personal information that makes
users identifiable in SNS with their actual personas that exist in the real world.
Information relating to digital identity can be divided into three tiers (Weik & Wahle
2010; Windley 2005) as shown in Figure 2.3. The first tier is called ‗my identity‘,
which contains persistent identity information such as a person‘s name, date of birth
and genealogical relations—information that is constant and unlikely to change
(Rowe 2010; Weik et al. 2010).
The second tier is called ‗shared identity‘, which contains attributes assigned to an
individual by other people, such as the social network of a person. Shared identity
contains information which is susceptible to change as a person makes friends with
different people and loses contact with others (Rowe 2010; Weik et al. 2010).
The third tier is called ‗abstracted identity‘ and contains identity information derived
My Identity
Shared Identity
Abstracted Identity
More Detailed
Information
Figure 2. 3 Three tiers of digital identity
(Source: Rowe (2010))
36
from groupings and demographics. For example, identifying a person by a
community of practice that they are involved in SNS. Identity information within this
tier is very likely to alter as a person‘s interests evolve over time. As Figure 2.3
illustrates, when users move up the tiers, the information describing the digital
identity of an individual becomes increasingly detailed and, therefore, deterministic
in the sense of uniquely identifying the person (Rowe 2010; Weik et al. 2010). Thus,
creating identities and sharing this personal information on SNS raises privacy
concerns for users of SNS.
2.5.1 Willingness to Share Digital Identity
Human behaviours differ across individuals, as each one will possess differentiated
intentions and interests as represented by their willingness to create and share
personal identity that is true or authentic, rather than that which is fictitious.
Behavioural intention is defined as ‗the degree to which person has formulated a
conscious plan to perform or not perform some specified future behaviour‘
(Warshaw & Davis 1985).
Our understanding of user behaviour is normally derived from studies within
marketing disciplines, but many studies that can be readily applied to online SNS
have also been conducted in relation to e-commerce. Some theories which are used
to study the effect of behaviour intentions by users upon their interactions within
SNS include the Theory of Reason Action (George 2002; Lo et al. 2010; Malhotra et
al. 2004; Shin 2010a), the Theory of Planned Behaviour (Ellison et al. 2007b; Lo
2010), Technology Acceptance Model, and Social Contract Theory (Fogel et al.
2009).
According to Bateman, Pike and Butler (2011), user behaviour is affected by
37
emotional responses that rely on external surroundings and internal individual
characteristics. In SNS, environmental characteristics involve creating groups,
communities, entertainment, sharing applications and research interests (Park, Kee &
Valenzuela 2009). Mori, Sugiyama and Matsuo (2005) suggest that individual
characteristics involve making friends, developing new relationships, and updating
personal identity and information. Moreover, DiMicco et al. (2008) suggested that
users with a high profile have more friends. However, Ybarra and Mitchell (2008)
argue that users who share more of their identity information may have different
intentions, for example, they may want to impress their co-workers, colleagues and
friends. Hence, they suggested that users have different intentions in terms of
willingness to share identity information in SNS. According to Ajzen (1991), users‘
willingness to provide information can be seen as a behavioural intention, which
according to Theory of Planned Behaviour (TPB) is a reliable predictor of actual
behaviour. Information disclosure involves submitting personal information to the
SNS. Regardless of whether or not the user chooses to, or is willing to share this
information with other online users, once submitted, the information is in the
possession of the SNS, and for the user this identifying information may be
unknowingly shared on the SNS. Finally, Lo (2010), Shin (2010b), and Hu and Ma
(2010) have studied the issues of privacy and trust relationship with users‘ different
attitudes and found that TPB is the most suitable method to measure attitudes
towards behaviour.
2.6 Gaps in the Literature
Academic literature on the subject indicates that information can be leaked through
SNS (Rosenblum 2007). This is because information disclosed through SNS opens
an opportunity for others to view the personal information and identity details for
38
that particular user (Molok, Chang & Ahmad 2010). Much research has been
undertaken in terms of privacy concerns and trust in e-commerce (Dwyer et al. 2007;
Lo 2010; Shin 2010b), but only a few studies have researched privacy concerns and
trust in the context of SNS. Internet knowledge and experience influence a user to
use the Internet for communication and to exchange information (Lo et al. 2010;
Park, Konana, Gu & Man Leung 2010). However, Fogel and Nehmad (2009) have
found that there is a risk of sharing information on the Internet due to privacy and
trust issues. They also suggested that the extent of users‘ knowledge of the Internet
knowledge might affect the degree to which they are willing to share identity
information. Furthermore, Lo and Riemenschneider (2010) found that Internet
literacy has a positive influence on sharing information, but they were unable to
determine whether or not privacy and trust played an important role in influencing a
user‘s willingness to share their digital identity on SNS. This study will focus on
understanding the impact of user privacy concerns and trust upon a user‘s
willingness to share their personal information on SNS.
2.7 Conclusion
This chapter has been concerned with the history and background of Social
Networking Sites. It also provides background information about Internet privacy
and user digital identity information sharing as it applies to the Internet. It was
further discussed how user privacy concerns and trust impact upon a user‘s
willingness to share information. However, despite privacy concerns, user
experience, knowledge and trust all playing important roles in governing the extent
of information sharing by users within SNS, there has been little empirical research
which has focused on which factors influence users‘ behaviour towards use of SNS
39
for sharing personal information. This research seeks to address those gaps in
knowledge in the field.
The next chapter explores theories related to behaviour, and the conceptual model
used will show the relationship between SNS experience, privacy concerns, trust and
their impact upon the user‘s willingness to share their digital identities. Finally,
hypotheses are developed to support this study.
40
Chapter III: Theoretical Support and Conceptual
Model
3.1 Introduction
This chapter describes in detail the concepts and theories drawn upon in support of
this study. Different theoretical bases can help explain different motivations and
intentions for users within Information Systems (IS), such as the Theory of Social
Capital, Social Exchange Theory, Theory of Weak Ties, Theory of Reason Action
(TRA), Technology Acceptance Models (TAM) and Theory of Planned Behaviour
(TPB). This chapter also describes the conceptual model design for this study, which
was developed from the literature review and supports the objective of this study.
This chapter also explores the research question that guides this study and the
hypotheses that arise out of it.
3.2 Related Social Theories
From previous study it has been established that a number of social theories have
been extensively used to explain why community members share their knowledge
with other users within the context of social networks. These theories suggest that
users expect either economic or social reward from their participation (Constant,
Sproull & Kiesler 1996; Nahapiet & Ghoshal 1998; Wasko & Faraj 2005). The
Theory of Weak Ties provides valuable insights into why people seek information
from virtual communities (Granovetter 1973; Granovetter 1983). Weak Ties refer to
relationships between people with little familial or occupational connection.
Consisting of members with diverse backgrounds, virtual communities can provide a
41
good platform within which people might diversify their sources of information.
Psychology-based theories, on the other hand, propose that individuals‘ decision
making-processes are often influenced by their perceived level of knowledge or
confidence (e.g., illusion of knowledge or overconfidence), leading them to make
decisions according to their prior beliefs, rather than by seeking advice from others
(Barber & Odean 2001). Such factors may mitigate members‘ intentions to seek
information from virtual communities, but cannot explain why they post information
about themselves. Information systems (IS) theories and models such as the Theory
of Reasoned Action (TRA) and Theory of Planned Behavior (TPB) can be extended
to explain how people are motivated to either share their knowledge or to seek
information from other members which, in turn, leads to different intentions and
behaviours in virtual communities (Lin 2006). In particular, the information adoption
model TPB that is extended from TRA posits that the usefulness of information,
quality of information, and trust toward information sources are important factors
that drive people to accept information from other sources (Bhattacherjee & Sanford
2006; Sussman & Siegal 2003).
This study draws upon the Theory of Reasoned Action (TRA) and Theory of Planned
Behaviour (TPB) to explain the phenomenon of users‘ willingness to disclose
personal information on SNS. The TRA (Ajzen et al. 1980) and its extension, the
TPB (Ajzen 1991), have been well adopted and applied by information systems
researchers for the last two decades in their examination of users‘ intentions to adopt
technologies (e.g. Davis, 1989). TPB postulates three determinants of behavioural
intentions: attitudes, subjective norms, and perceived behavioural control; however,
underlying these determinants are the beliefs that are relevant to the behaviour
42
(Ajzen 1991). In this study, the focus is on investigating how individuals‘ beliefs
about Internet privacy and their trust in different entities influence their willingness
to provide personal information on SNS. In general terms, if an individual has
positive beliefs regarding a certain behaviour, then he or she would be more willing
to engage in that particular behaviour (Ajzen 1991).
3.2.1 Theory of Reasoned Action (TRA)
The Theory of Reasoned Action (TRA) describes ‗intention‘ as the best predictor of
whether or not behaviour is performed. According to TRA, direct determinants of
behavioural intention are the pre-existing attitude towards the behaviour and the
subjective norm associated with the behaviour. Attitude refers to personal beliefs
about the positive or negative value associated with a particular behaviour and its
outcomes. The term subjective norm refers to a person‘s positive or negative value
associated with a given behaviour. One‘s attitude depends on whether or not the
behaviour that is being considered is accepted by important referent individuals and
their motivation to comply with those referents. Ultimately, interventions can be
designed to change behavioural intentions by affecting one‘s attitude and subjective
norm about a particular behaviour in order to promote that specific behaviour in a
person.
TRA was drawn from social psychology. It is one of the fundamental theories of
human behaviour and has been used to predict behaviour in a broad range of
dimensions. Davis (1989) originally applied TRA to individual acceptance of
technology and found that the variance explained was largely consistent with studies
that had employed TRA in the context of other behaviour. Researchers in the domain
43
of information systems use this theory to understand the adoption of IT innovation
(Han, 2003). TRA has been employed in education, automation in manufacturing and
in Internet banking. Even though there is evidence that this theory can be used to
understand the adoption behaviour for new technologies, there is limited evidence
that this can be used to understand the determinants in understanding the human
behavioural intention to use their own sensitive information to share with other
systems.
3.2.2 Technology Acceptance Model (TAM)
The Technology Acceptance Model (TAM) (Davis 1989; Davis 1993) is an
adaptation of the Theory of Reasoned Action (TRA) (Fishbein & Ajzen 1975) which
specifies two beliefs—perceived usefulness and perceived ease of use—as
determinants of attitude towards usage intentions and usage (Davis 1989). Usage
intentions are, in turn, the sole direct determinants of usage (Taylor & Todd 1995b).
Introducing intentions as a mediating variable in the model is important for both
substantive and pragmatic reasons. Substantively, the formation of an intention to
carry out behaviour is thought to be a necessary precursor to behaviour (Fishbein et
al. 1975). Pragmatically, the inclusion of intention is found to increase the predictive
power of models such as TAM and TRA, when compared to models that do not
include intention (Fishbein et al. 1975; Taylor et al. 1995b).
3.2.3 Theory of Planned Behaviour (TPB)
In the development of theory, Theory of Planned Behaviour (TPB) is derived from
the Theory of Reasoned Action (TRA). TPB mentions attitude, subjective norm and
44
perceived behaviour control intention. In this context, intention indicates people‘s
desired effort to conduct an activity. Furthermore, attitude is an individual preference
towards the object in question (Yu & Wu 2007). According to Crespo, Herrero and
Bosque (2008) subjective norms reflect the degree of people‘s affection towards an
object or behaviour, due to the perception of a significant referent/s. Moreover, they
argue that perceived behavioural control reflects the perception of the availability of
resources and opportunities for behaviour development.
Associated with intention to use SNS to share information using TPB, a study carried
out by Peluchette and Karl (2008) found that attitudes and subjective norms
positively affect an individual‘s intention to share information, whereas perceived
behavioural control does not support this intention. This study includes personal
innovativeness as a moderating effect. Another study conducted by Lin (2006) found
that attitude and perceived behavioural control does positively affect willingness to
share information, while subjective norms do not support this willingness. These
studies were conducted with participants of varying backgrounds and antecedents.
TPB draws upon constructs taken from literature relating to characteristics, and more
completely to account for conditions where individuals do not have complete control
over their behaviour. The TPB asserts that behaviour is a direct function of
behavioural intention and perceived behavioural control, and that behavioural
intention is formed by one‘s attitude which, in turn, reflects a feeling of
favourableness or unfavourableness towards performing a behaviour; and a
subjective norm which reflects the perception that significant referents desire the
individual to perform or not perform a behaviour; as well as perceived behavioural
45
control, which reflects an individual‘s perceptions of internal and external constraints
on behaviour (Ajzen 1991).
Moreover, TPB contains an additional determinant, perceived behavioural control, to
accommodate deficiency control and resources for a particular behaviour—which
can be deliberate and planned. TPB is considered to be generic as well as assuming
that individuals will use the information available logically with rational decision
making. This assumption has been used to understand and explain behaviour across a
wide range of domains, such as marketing and consumer behaviour and leisure
behaviour. All this previous evidence asserts that this theory can be used to
understand the adoption of human behaviour to control attitudes towards performing
behaviour.
3.3 Research Question (RQ)
To what extent do privacy concerns and trust impact users’ willingness to share
digital identities within social networking sites?
To answer the main RQ the following sub-questions will be addressed:
RQ1 Does a user‘s experience of SNS have a positive impact upon their willingness
to share digital identity?
RQ2 To what extent does users‘ privacy concerns impact on trust as a determinant in
their willingness to share digital identities on SNS?
RQ3 To what extent does trust influence users to share their digital identities in
SNS?
46
RQ4 To what extent do privacy concerns influence users to share digital identities in
SNS?
3.4 Conceptual Model
The conceptual model is developed for this research using the Theory of Planned
Behaviour by Ajzen (1991), and concepts of privacy and trust developed from
literature review and previous papers by Lo and Riemenschneider (2010), Dinev and
Hart (2006), and Malhotra, Kim and Agarwal (2004).
This model postulates that privacy concerns, trust and SNS experience do affect
H5
Willingness
to share
Digital
Identites
Privacy
Concerns
SNS
Experience
Trust
H2
H4
H3
H1
Figure 3. 1 Conceptual model - Key factors of SNS that impact on
willingness to share digital identity
(Source: The Author)
47
users‘ willingness to share their digital identity. Privacy concerns and trust play an
important role in influencing behaviour, therefore, this model applies the Theory of
Planned Behaviour developed by Ajzen (1991). According to TPB (Ajzen 1991, p.
179), ‗people‘s willingness to provide information can be seen as an individual
behaviour driven by behavioural intensions where behavioural intentions are a
function of an individual‘s attitude toward the behaviour, the subjective norms
surrounding the performance of the behaviour and the individual‘s perception of the
ease with which the behaviour can be performed (behavioural control)‘. In this model
privacy concerns and trust perceptions are behavioural controls of individuals that
will determine a user‘s belief and willingness to share information. Thus, the model
will test the effects of privacy, trust and SNS experiences upon users‘ willingness to
share their digital identity.
In this study, the author was interested to know how privacy concerns and trust
impacted on users‘ willingness to share personal information on SNS. People‘s
willingness to provide information can be seen as a behavioural intention which,
according to TRA/TPB, is a reliable predictor of actual behaviour (Ajzen 1991).
Information disclosure involves submitting personal information to the SNS.
Regardless of whether or not the user chooses to or is willing to share this
information with other online users, once submitted, the information is in the
possession of the SNS. While people can easily (and do) provide fictitious
information to social networking sites, the focus of this study is on investigating why
people are willing to provide authentic personal information. For the purposes of this
study, personal information broadly encompasses any information that can help trace
and confirm one‘s identity, such as name, birth date, address, phone number,
48
photograph, or the location of their home town. Prior studies about information
sharing on SNS suggest that trust in a SNS is a driving force that increases a person‘s
willingness to share information on that SNS (Dwyer et al. 2007; Fogel et al. 2009).
The Theory of Planned Behaviour suggests that beliefs related to a behaviour are
prevailing determinants of a person‘s behavioural intentions (Ajzen 1991). As such,
in the context of this study, a person‘s beliefs about their knowledge and experience
are determinants of his or her intentions to share personal information on a SNS. On
the one hand, lower degrees of Internet experience may elevate the user‘s concerns
about privacy, because although the individual may be vaguely aware of potential
dangers, he or she may not know how to manage them. On the other hand, higher
degrees of Internet experience can likewise raise concerns about privacy, because the
individual is more aware of the dangers. However, the more literate individual may
have the knowledge to attempt to minimise the dangers by installing and updating
operating system and browser security patches and fixes, anti-spyware software and
alerts, and other prevention utilities. Due to this perception of control (the ability to
try to minimise potential dangers), individuals with higher SNS knowledge and
experiences are expected to be less concerned about their Internet privacy than
individuals with lower levels of knowledge. For example, in a study of the
behavioural intention to conduct online e-commerce transactions, Dinev and Hart
(2005-6) found that higher Internet literacy was a positive intention on willingness to
carry out transactions online. Hence, the first hypothesis is derived:
H1: Users with high levels of SNS experience will be more willing to share digital
identities.
49
Drawing upon the TRA/TPB (Ajzen 1991; Ajzen et al. 1980), it is argued here that a
person‘s privacy concerns and trust beliefs will tend to significantly influence his or
her willingness to provide personal information to a SNS. The TPB postulates three
determinants of behavioural intention: attitudes, subjective norms, and perceived
behavioural control; however, underlying these determinants are the beliefs relevant
to the behaviour (Ajzen 1991). In general, if an individual evaluates beliefs regarding
a behaviour in a positive light, then he or she would be more willing to perform the
particular behaviour (Ajzen 1991). Conversely, a person who evaluates beliefs
regarding a particular behaviour negatively will be less likely to perform the
behaviour. In studies of privacy concerns and trust in the e-commerce literature,
perceived risk was found to negatively influence people‘s willingness to provide
personal information to transact online (Malhotra et al. 2004; Van Slyke, Shim,
Johnson & Jiang 2006). Therefore, based on the above studies and literature review,
the following hypothesis is used to test the proposed model:
H2: Users with higher levels of privacy concerns have a lower level of trust in SNS.
Similar studies conducted by Malhotra et al (2004) and Liu et al (2005) about trust in
online transactions (e-commerce) concluded that trust was found to positively
influence the user‘s intention to use e-commerce. Hence, based on the above studies
and literature review, the following hypothesis was used to test the proposed model
between trust and willingness to share digital identities.
H3: Users with higher levels of trust will be more willing to share digital identities
on SNS.
50
As discussed previously, the TPB contends that behavioural intentions are
antecedents to the specific behaviours of an individual. More specifically, an
individual‘s attitudes and perceptions will influence that individual‘s actions when he
or she believes that certain behaviour will be linked to a specific outcome. Further,
subjective norms and social pressures about whether or not to engage in a particular
behaviour influences behavioural intentions, as determined by an individual‘s
positive or negative evaluation of it (Liu et al. 2005). Based on the same logic, a
user‘s perception of and attitudes about privacy should influence his or her attitudes
toward their willingness to share information; and, in turn, shape his or her
behavioural intentions about their participation in a SNS. Hence, based on the above
studies and literature review, the following hypothesis is used to test the proposed
model between privacy concerns and willingness to share digital identities.
H4: Users with higher levels of privacy concerns will be less willing to share digital
identities.
3.5 Conclusion
This chapter reviewed and discussed several traditional theories used in information
systems from the last two decades. As the project is related to human behaviour with
technology, this study adopts and uses the Theory of Planned Behaviour (TPB),
which is derived from the Theory of Reasoned Action (TRA) developed by Ajzen in
1991. Building on the literature review as presented in the previous chapter, a
research model was developed. This research model can contribute to a more
comprehensive understanding of the relationship between concerns about trust and
51
privacy as they relate to a user‘s willingness to share digital identities and SNS
experiences. Together, it has the potential to offer a richer explanation of the impact
of this variable on sharing information on SNS.
The author further identified and reviewed the independent and dependent variables
in the model and used relevant theories to derive the variables and explain their
hypothesised relationships with the dependent variables. The next chapter describes
the methodology; and operationalises the constructs in order to test the hypotheses as
depicted in the model.
52
Chapter IV: Research Design and Methodology
4.1 Introduction
There are different types of research such as exploratory, descriptive, analytical,
predictive, quantitative, qualitative, deductive, inductive, applied and basic research
(Collis & Hussey 2009). The main unifying theme that unites all types of research is
the need for researchers to focus their efforts on answering two major significant
questions (Kripanont 2007).
These questions are, firstly, what methodologies and methods will be used in the
research? Secondly, how do they justify this choice and use of these methodologies
and methods? Justification of the researcher‘s choices and use of particular
methodology is something that underpins assumptions about reality that they bring to
their work (Crotty 1998).
The research methodology and methods for this study were chosen in order to
successfully achieve the essential research objectives. The justification of choices
and uses will be presented in this chapter, as detailed in the explanation of the chosen
research methodology and design of this study. The study materials contained in this
chapter re-state the research objective, and then specify the research design, survey
population, sampling procedures, instrumentation, data collection and data analysis.
This study is guided by the following question: To what extent do privacy concerns
and trust influence users‘ willingness to share their digital identities on social
networking sites? Within this chapter the development of the relevant instrument for
addressing the fundamental research question posed by this thesis (in this case by
means of a survey) is discussed.
53
4.2 Objective
The broad objective of this study is to show the relationship between privacy
concerns, levels of trust of SNS users and their willingness to share their identities
with the other SNS users. To reach the goal in a meaningful way, the author has set
the research questions as follows:
i. Does prior experience of SNS have a positive impact on the users‘ willingness
to share aspects of their digital identities?
ii. To what extent does the users‘ concern about privacy impact upon their level
of trust that in turn guides their willingness to share digital identities on SNS?
iii. To what extent does trust influence users‘ decisions to share digital identities
on SNS?
iv. To what extent do privacy concerns influence users to share their digital
identities on SNS?
4.3 Research Design
The development and refinement of a research design involves a series of rational
choices. The research design is the process aimed at designing the research study in
such a way that the essential data can be gathered and analysed to arrive at a solution
(Sekaran 2003).
This study uses an explanatory design so that the findings of this research are based
upon quantitative study (Creswell, Hanson, Clark & Morales 2007). Bryman and
Bell (2007) demonstrate—as depicted in Figure 4.1—the ideal path to follow when
organising research by a quantitative method. However, they also clearly state that
this work process is an ideal way that is rarely followed in any research in practical
54
terms. Literally, the actual process followed in the research may include going
backwards and forwards between these steps (Bryman et al. 2007; Liong & Mejstad
2010). Instead, they argue that the illustration should serve as a presentation of the
most important steps to incorporate when performing quantitative research.
Figure 4.1 The process of quantitative research
(Source :Bryman and Bell (2007))
For this study, the work process flow has not followed the ideal path outlined.
Instead, this study does go back and forth between the different steps on its progress
towards completion. In doing this, this researcher makes sure that all the steps have
in fact been followed, and this process model thus helped the researcher to ensure
55
that all the components were included and connected to each other, and the
modifications subsequently made to the research design are shown below in Figure
4.2.
Figure 4.2 The flow chart of research design
(Source: The Author)
4.4 Research Philosophy and Paradigm
An appropriate research paradigm is an essential concept for any research study.
Therefore, a research paradigm can be viewed as a world-view for understanding the
56
complexities of the real world, or assumptions relating to a world which is shared by
a society of researchers exploring that world. The research used as the guiding
paradigm was positivist, in which measurement and quantification were emphasised
in the pursuit of objective knowledge (Seale 1999). This is a scientific approach in
which the researcher works logically, and in which the data collection and data
analysis aspects of the research are seen as highly important (Creswell & Clark
2007). For the purposes of this research, online community success was a value
judgment made by an individual. This perspective is consistent with Wenger‘s
(2005) view that although social networking sites may be designed for communities,
individual members were the ones who ultimately experienced the technology as
they engaged with the communities. It also resonates with Seddon‘s (1997)
observation that different people using the same system may draw very different
conclusions about the success of the system.
By using quantitative research methods, this study was able to perform statistical
testing, thereby allowing for higher levels of generalisability than what would be
possible using qualitative research methods (Seale 1999). However, this study agrees
with criticism of the positivist paradigm, most notably stemming from researchers
with a post-positivist perspective (Fischer 1998). According to this criticism, the
positivist paradigm is described by post-positivists as naïve, as there is no one
absolute truth that can be unravelled by relying on positivist research methods.
According to Fischer (1998), rather, reality is constructed and is dependent on the
perspective of the individual. But even if this criticism is pertinent, especially to
researchers within the social science area, this research argues that choice of
paradigm and research method for this study has been done in line with post-
57
positivist research‘s emphasis on multiple methods of inquiry (Fischer 1998).
4.5 Research Approach
The specific approach of this study was to build upon previous research regarding
privacy concerns and trust related to SNS. Moreover, from its literature review and
theoretical background, this research was carried out with an approach that was
deductive in overall terms (Bryman et al. 2007). According to Bryman and Bell
(2007) the deductive approach is one in which hypotheses are commonly stated,
based on previous findings and then tested using statistical methods. The study‘s
conclusions are then based on the rejection or non-rejection of the stated hypotheses.
However, Bryman and Bell (2007) note that a significant portion of quantitative
research does not include a statement of hypotheses in the first place; that it is a
feature most pervasive in experimental studies. This research was carried out in
accordance with research processes based on the concepts of the hypothetico-
deductive method (Sekaran 2003).
The following are the steps that were conducted throughout this study project:
1) Observation (which was conducted, but it was not used as methodology).
2) The use of a semi-online survey to collect data to ascertain what is happening and
why, the collected data will be useful for study; and the questionnaire can be
amended if necessary.
3) A broader survey of the literature was conducted so that the researcher might
anticipate a myriad of scenarios that may arise during the study, and determine
how to overcome each situation. This information gives additional insights into
58
the various possibilities that might eventuate, and helps confirm that these
variables were good predictors for privacy concerns for users and the effect of
trust upon user s‘ willingness to share their digital identities.
4) Theory formulation is a step in developing a theory incorporating all the relevant
factors contributing to usage behaviour and behavioural intention of academics to
use the Internet and SNS. It was an attempt to integrate all the information in a
logical manner, involved a collection of theories and models from the literature to
help conceptualise and test the reasons for the problems. In other words, it
explained the research questions and hypotheses, and helped in the identification
and labelling of variables (Hussey & Hussey 1997).
5) Hypothesising
This step was used to generate various hypotheses that would be tested to examine
whether the theory formulated was valid or invalid.
6) Data collection
A questionnaire was developed, based on the previous literature and research; to
determine use, privacy concerns and trust as factors that might influence SNS
users‘ willingness to share their digital identity. This was then used as a survey
tool to collect data.
7) Data analysis
The data that was collected through the online survey questionnaire was analysed
to see what factors influenced users to share digital identities on SNS. Other
information about the user, such as demographic data and their interest in and
knowledge of social networking sites, can be obtained at this stage.
8) Deduction
This is the process of concluding the results by interpreting the meaning of the
59
results of data analysis.
4.6 Questionnaire Design
The online survey used in this study was created by the researcher using ‗Qualtrics‘
and was physically located on secure servers of the Qualtrics company. Qualtrics was
used to collect and store the data for analysis. The application was configured to
support anonymous participation, thereby assuring anonymity, confidentiality and
privacy throughout the study, and making it impossible for the researcher to make
links between survey responses and specific responders. The online form of the
survey also allowed the participants to complete the survey at any time and in a
location convenient to them. In addition, the survey was configured so that
participants could exit and start the survey at any time and return at a later time/date
to recommence the survey.
The overall design of this research questionnaire was highly inspired by Dinev and
Hurt (2006), Malhotra, Kim and Agarwal (2004), McKnight et al. (2003), Lo and
Riemenschneider (2010) and Fogel and Nehmad‘s (2009) design. Similarly, in order
for this research to also answer its own research questions, the questionnaire had to
be divided into several parts. The first part revolved around demographic information
and general knowledge of social networking sites. The second part concerned the
general attitude of respondents towards privacy concerns. The third part related to the
trust expressed by respondents about social networking sites; and the fourth and final
part concerned the user‘s willingness to share digital identities on SNS.
According to Sekaran (2003), in order to minimise bias in the results of the research,
60
researchers need to concentrate mainly on three areas when designing a
questionnaire:
i) The wording of the questions;
ii) Planning how the variables will be categorised, scaled and
coded after receipt of the responses, and
iii) The general layout content of the questionnaire design.
Overall, table 4.1 presents the questions asked in this research questionnaire and the
scientific background pertaining to them. As such, it was up to the personal judgment
of the researcher to construct relevant measurements, however, the item used to
measure the research model (items used in the questionnaire) was developed from
the items already used by Dinev and Hurt (2006), Malhotra, Kim and Agarwal
(2004), McKnight et al. (2003), Lo and Riemenschneider (2010) and Fogel and
Nehmad (2009). In addition, the scientific background mentioned in the table refers
to researchers who have used similar questions to assess attitudes within the realm of
SNS experiences, privacy concerns, trust and willingness to share digital identities on
SNS.
The design process for the questionnaire for this project took almost 3 months (May
2011 to July 2011) prior to the pilot online survey being undertaken. This was
because the researcher was aware that in designing the questionnaire it was important
to proceed with caution, keeping in mind the rationale for why the research was
being done (Veal 2005). Thus, the researcher acknowledged that the aim in designing
the questionnaire was to achieve the research objectives (see Chapter I), with
consideration for meeting the basic criteria of relevance and accuracy (Zikmund,
61
Griffin, Babin & Carr 2009).
The questions were structured and separated into five sections (see Table 4.1 and
Appendix D). Some sections such as privacy concerns, trust and willingness to share
digital identity used a 7-point Likert scale because this method it is extremely
popular for measuring attitudes, and is relatively simple to administer. With the
Likert scale, respondents indicate their attitudes by checking how strongly they agree
or disagree with carefully constructed statements that range from very positive to
very negative toward the attitudinal object. In this study, seven alternative
measurement scales were used which were adopted from previous well-known
studies such as Dinev and Hurt (2006), Malhotra, Kim and Agarwal (2004),
McKnight et al. (2003), Lo and Riemenschneider (2010) and Fogel and Nehmad
(2009). This scale ranges from strongly disagree = 1, disagree = 2, somewhat
disagree = 3, neither agree nor disagree = 4, somewhat agree = 5, agree = 6, and
strongly agree = 7. A brief summary of the use of scales and measurements follows.
Section A focused on general knowledge, experience and use of the SNS as it related
to the demographic data of the users. This section contained 16 questions, and of the
16, five questions were open-ended and the rest of the questions were closed. It
comprised six questions that are established as ordinal scales, such as ‗How many
friends do you have in your Social Network accounts on average?‘ and ten questions
with nominal scales. The design of this section was based on previous studies.
Section B was an important section focused on respondents‘ privacy concerns about
using SNS. It was established as a 7-point Likert scale. It comprised five items based
62
on the previous study such as ‗I am concerned that the information I submit on the
Internet could be misused‘, (see Table 4.1 for full details).
Section C was also an important section used for establishing the trust concerns of
user. It was established as a 7-point Likert scale. Measurements were developed
from previous studies by McKnight et al. (2002), Dinev and Hart (2006), Malhotra,
Kim and Agarwal (2004) and including statements such as, ‗SNS would tell the truth
and fulfil its promises related to the personal information provided by me‘.
Section D focused on investigating the SNS user‘s intention to willingly share digital
identity information. This was established as a 7-point Likert scale.
Table 4-1 Questionnaire items and variable coding
Question/Statement Scientific Background Variable
Abbreviation
Demographic
In what age range are you? Author develop
What is your occupation? Author develop
What is your gender?
Fogel and Nehmad (2009); Hoy and
Milne (2010); Manago, et al. (2008);
Lewis, et al., (2008)
What is your marital status? Author develop
What is your highest education level? Author develop
Why do you use Social Networking? Author develop
I am using Social Networking Site Fogel and Nehmad (2009); Dwyer, et al., (2007)
I know Social Networking Site from Author develop
How many Social Networking accounts you have? Fogel and Nehmad (2009); Tufekci (2008)
How many friends do you have in your Social Network accounts in average
Fogel and Nehmad (2009); Lewis, et al., (2008); Ellison et al., (2007)
Has the usage of social networking made a positive
impact in your social life?
Author develop
What is your country of origin? Please specify Author develop
SNS Experience and Knowledge
I am using Internet since Fogel and Nehmad (2009) SNS1
I am using Social Networking Site since Fogel and Nehmad (2009) SNS2
Did you read the Terms of use/ Privacy policy while creating the Social Networking Site account?
Author develop SNS3
Did you modify privacy settings from default settings
after creating the account? Gross and Acquisti (2005); SNS4
Privacy Concerns
I am concerned that the information I submit onto social
networking sites could be misused.
Dinev and Hart (2006), Malhotra, Kim
and Agarwal (2004) PC1
63
Question/Statement Scientific Background Variable
Abbreviation
I am concerned that a person can find private information
about me on the Internet.
Dinev and Hart (2006), Malhotra, Kim and Agarwal (2004)
PC2
I am concerned about submitting information on the Internet because of what others might do with it.
Dinev and Hart (2006), Malhotra, Kim
and Agarwal (2004) PC3
I am concerned about submitting information on the Internet because it could be used in a way I did not
foresee
Dinev and Hart (2006), Malhotra, Kim
and Agarwal (2004) PC4
I am concerned that online companies are collecting too
much personal information about me
Dinev and Hart (2006), Malhotra, Kim
and Agarwal (2004) PC5
Trust
I believe that SNS would act in my best interest when dealing with my personal information.
McKnight et al. (2002), Dinev and Hart (2006), Malhotra, Kim and Agarwal
(2004
T1
SNS is interested in protecting my personal information according to the preferences I specify.
McKnight et al. (2002), Dinev and Hart (2006), Malhotra, Kim and Agarwal
(2004
T2
SNS would tell the truth and fulfill its promises related to
the personal information provided by me.
McKnight et al. (2002), Dinev and Hart
(2006), Malhotra, Kim and Agarwal (2004
T3
SNS is sincere and genuine in managing my personal
information.
McKnight et al. (2002), Dinev and Hart
(2006), Malhotra, Kim and Agarwal (2004
T4
SNS handles personal information submitted by users in
a competent fashion.
McKnight et al. (2002), Dinev and Hart
(2006), Malhotra, Kim and Agarwal (2004
T5
SNS performs its role of managing my personal
information according to my privacy settings very well.
McKnight et al. (2002), Dinev and Hart
(2006), Malhotra, Kim and Agarwal
(2004
T6
I believe that if I allowed my SNS friends to view my
personal information, they would act in my best interest
when dealing with this information.
McKnight et al. (2002), Bhattacherjee, A. (2002), Dinev and Hart (2006)
T7
My SNS friends would not use my personal information
opportunistically.
McKnight et al. (2002), Bhattacherjee,
A. (2002), Dinev and Hart (2006)
T8
I would characterize my SNS friends as honest in
handling my personal information.
McKnight et al. (2002), Bhattacherjee,
A. (2002), Dinev and Hart (2006)
T9
My SNS friends are sincere and genuine in dealing with my personal information.
McKnight et al. (2002), Bhattacherjee, A. (2002), Dinev and Hart (2006)
T10
My SNS friends have the skills and expertise to handle my personal information carefully.
McKnight et al. (2002), Bhattacherjee, A. (2002), Dinev and Hart (2006)
T11
Willingness to share digital identities
I am willing to share my real name.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI1
I am willing to share my real date of birth.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI2
I am willing to share my real hometown address.
Fogel and Nehmad (2009); Gross and Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010
DI3
I am willing to share my real email address.
Fogel and Nehmad (2009); Gross and Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI4
I am willing to share my real home phone number.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al., (2009); Lo and Riemenschneider (2010)
DI5
I am willing to share my real mobile phone number.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al., (2009); Lo and Riemenschneider (2010)
DI6
I am willing to share my real photograph.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI7
I am willing to share the name of my real high school(s),
I have attended.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al., DI8
64
Question/Statement Scientific Background Variable
Abbreviation
(2009); Lo and Riemenschneider (2010)
I am willing to share the name of my real college(s)
attended.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al., (2009); Lo and Riemenschneider (2010)
DI9
I am willing to share my real name of employer.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al., (2009); Lo and Riemenschneider (2010)
DI10
I am willing to share my real interests.
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI11
I am willing to share my real personality
Fogel and Nehmad (2009); Gross and
Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI12
I am willing to share my real gender Fogel and Nehmad (2009); Gross and Acquisti (2005); Christofides, et al.,
(2009); Lo and Riemenschneider (2010)
DI13
4.7 Data Collection and Sample Size
In order to examine SNS users‘ privacy concerns and trust and behaviour, and
willingness to share their digital identity in SNS, the population of this research (i.e.
the entire group of people that the researchers‘ desires to examine) (Sekaran 2003)
are all SNS users. Consequently, if this study were to use a probabilistic sampling
method, it would require the construction of a selection procedure in which all SNS
users would have the same likelihood of being selected to participate in this study
(Anderson, Sweeney, Williams, Freeman & Shoesmith 2010). Regardless of how the
probabilistic sampling method would be performed in practice (simple, stratified or
cluster), this requirement was simply not be possible to fulfil due to the limited time
available for this Masters level study.
To collect data, an online survey link was posted on different SNS forums, and
requests were made for users to participate in the survey. This study was careful in
choosing a Facebook forum where the visitors would, to a significant extent, include
persons displaying the same diversity in attitudes as the general public. The sample
population of this study is taken from Facebook users where, according to Facebook
(2011), they have more than 500 million users; however, for this study purpose data
is collected from Facebook Forum users where the number of registered users is
65
approximately 17,000 and a list of the users is publicly available in the forum and the
total number of participants in the online survey were 155 users. Facebook Forum is
a community forum where users are able to share, discuss and comment on anything
related to Facebook and other SNS, such as a new SNS feature, its advantages and
disadvantages, an application, games, new privacy settings, scams, photos, videos,
events, and numerous other general topics where users can request help. There are
also forum contests, general chat areas and an interactive Chatbox, where online
members gather to chat together in real-time (Forum 2011). Facebook Forum was
selected for this study because it is a subset of Facebook and it will improve the
credibility of the results because participation includes users from different
backgrounds including application developers, professionals, students and social
workers, who represent much broad population of SNS users.
As this research collected data on the number of people who have visited the
questionnaire, as well as the number of people who have completed it, an assessment
can be made regarding this type of non-response error. It was found that 191 people
visited the survey and out of those 191 only 161 people attempted the survey. Of
these 161 participants only 155 completed the entire questionnaire, which
corresponds to a mean survey completion rate of approximately 96.27%. This value
is reflective of good questionnaire design.
Another data collection method that could have been used is posting links to
questionnaires in different SNS forums. Even if that had increased the sample size,
the detrimental effects on the generalisability of these research findings would have
been too significant. Overall, this study argues that the data collection method
66
actually employed by the researcher ensured that the researcher received a sample
that was more diverse than what would be the case using any other available
methods.
4.8 Ethical Considerations
In any research study, ethical clearance is important, and is mandatory if the research
involves humans. This study directly involved people through the process of the
online survey instrument. Therefore, procedures were followed to gain ethical
clearance from the USQ Ethics Committee. At the same time, participants in the
online survey were clearly notified about their voluntary participation, the
confidentiality of the data and the participants‘ identities. Furthermore, participants
in this research were informed about their right to privacy and their option of
discontinuing their participation in the study at any time. In this research, an
informed consent was implied to the participations. Anonymity of the respondents
was guaranteed: there was no entry in the questionnaire to identify a specific
respondent, so it is impossible for the researcher to identify any individual response.
Furthermore, all the data gathered in this study were kept secure and confidential,
according to USQ regulations.
Data and information gathered in this study were stored in digital format on a secure
USQ server. It was also clear to the participants that under no circumstances would
the identity of any individual or group of individuals be released in any publications
that may eventuate from this study.
67
4.9 Data Analysis
This section presents an analysis of the data that was collected from the online
surveys. Non-parametric techniques are ideal for use when the collected data are
measured on nominal and ordinal scales. They are also useful when the samples are
relatively small and when data do not meet stringent assumptions of the parametric
techniques (Pallant 2011). With this in mind, Statistical Package (SPSS 19.0) was
used to analyse the data. Following the analysis of demographic data, Exploratory
Factor Analysis was then conducted to check if the proposed factor structures were
indeed consistent with the actual data. The various loadings are shown in a number
of tables in Chapter 5. Next, Confirmatory Factor Analysis (CFA) was conducted to
check the reliability and validity of the measurement model. This measurement
model was estimated using AMOS 19.0. A correlation study investigated the
relationship between independent and dependent variables using a Structural
Equation Modelling (SEM), a casual modelling statistical tool.
4.9.1 Descriptive Statistics Analysis
The purpose of conducting descriptive statistical analysis is to illustrate the
characteristics of the constructs associated with frequencies, the mean values, and
standard deviation of each variable. Descriptive statistics is also designed to provide
information about the distribution of the variables (Cooper & Schindler 2003).
4.9.2 Reliability
Testing the veracity of the data is done through testing the reliability and validity of
the measures. According to Veal (2005), reliability is the extent to which research
68
findings would be the same if the research were to be repeated at a later date, or with
a different sample of subjects. In other words, the reliability of a measure indicates
the extent to which the measure is without bias (error free) and, hence, offers
consistent measurement across time and across the various items in the instrument. It
helps to assess the goodness of measure, and indicates accuracy in measurement
(Sekaran 2003).
The most common and popular way to check reliability is by using Cronbach‘s alpha
(Cronbach 1951; Peter 1979; Sekaran 2003) or the mean inter-item correlation
between the items (Pallant 2011). This is a test of the consistency of respondents‘
answers to all the items in a measure. To the degree that items are independent
measures of the same concept, they will be correlated with one another (Sekaran
2003). The Cronbach‘s Alpha for each construct was calculated to check the
reliability of the scales. The calculation also provided a baseline for the analysis of
internal consistencies. The Cronbach‘s Alpha for each construct is displayed in Table
4.2. All constructs were found to have a Cronbach‘s Alpha higher than 0.6.
According to Sekaran (2003), reliabilities less than 0.6 are considered to be poor;
those in the 0.7 range are acceptable; and those over 0.8 good. The closer the
reliability coefficient gets to 1.0, the better. In other words, the generally agreed upon
lower limit for Cronbach‘s alpha is 0.70, but this may decrease to 0.60 in exploratory
research (Robinson, Shaverm & Wrightsman 1991). The results suggest that the
items consistently measured the constructs and were suitable inclusions in the final
scales (Nunnally 1994). Moreover, the items and measurement are adopted from the
previous studies so that they are reliable for this study.
69
4.9.3 Validity
Validity is defined as the extent to which the data collected truly reflects the
phenomenon being studied. Usually, business research faces difficulties about
validity, specifically in the measurement of attitudes and behaviour since there are
always doubts about the true meanings of responses made in surveys, interviews, and
self-reporting of behavior (Malhotra 2008; Veal 2005).
Sekaran (2003) suggests several types of validity tests for testing the integrity of
measures including content validity, criterion-related validity, and construct validity.
4.11.3.2 Construct Validity
The construct validity that was used in this research testified as to how well the
results obtained from the use of the measure fit the theories around which the test
was designed. In other words, construct validity testified that the instrument did tap
the concept as theorised. Construct validity can be established through (1) correlation
analysis (convergent and discriminant validity), (2) factor analysis, and (3) the multi-
trait, multi-method method matrix of correlations. Others suggest the three most
widely-accepted forms of validity are convergent, discriminant, and nomological
validity (Biesanz & West 2004; Campbell & Fiske 1959).
Convergent validity is synonymous with criterion validity (Zikmund et al. 2009) and
with correlation analysis, and is one way of establishing construct validity for this
research. It indicates that items that are indicators of a specific construct should
converge or share a high proportion of variance in common (Hair Jr et al. 2006). In
other words, it assesses the degree to which two measures of the same concept are
70
correlated, with high correlation indicating that the scale is measuring its intended
concept. Thus, reliability is also an indicator of convergent validity (Hair Jr et al.
2006).
According to rules of thumb, it has been suggested that item-to-total correlations
exceed 0.50 and the inter-item correlations exceed 0.30 (Robinson et al. 1991).
Cohen (1988) suggests correlation (r) = 0.10 to 0.29 (small correlation: both positive
and negative correlation), r = 0.30 to 0.49 (medium correlation), and r = 0.50 to 1.00
(large correlation). As results of the inter-item correlation values of the indicators in
each construct being in both medium and high levels (higher than 0.30, and most of
them higher than 0.50) (except some inter-items correlation values in usage
behaviour), and the item-total correlation values were also in higher levels (higher
than 0.50) (except some item-total correlation values in usage behaviour), these
indicated the convergent validity of the instrument.
Because of the reliability of results with high coefficient alpha, and the correlation
values of the questionnaire and the results of the convergent validity of this pilot
study, a minor change was made to the wording of the questionnaire after the pilot
study. The instrument was developed and designed based upon the theoretical
literature survey and adopted from previous studies. Thus, the measures of the
instrument provided adequate coverage of the concepts; and the instrument has clear
and understandable questions. Consequently, the instrument was reliable, valid when
considering content validity, construct validity and theoretical validity; and was
ready to be used in the main survey.
71
4.9.4 Factor Analysis
Factor analysis is used to explore the underlying pattern or relationship for a large
number of variables and to determine whether the information can be condensed or
summarised in a smaller set of factors or components (Hair Jr et al. 2006). Construct
validity will be examined via Confirmatory Factor Analysis (CFA). CFA is to be
used when the research knows about the number of the factors, as well as which
variables load on the specific factors (Hair Jr et al. 2006; Liao, Chen & Yen 2007).
4.9.5 Structural Equation Modelling (SEM)
The main objective of this SEM analysis was to generate a model that best described
the effect of privacy concerns and trust as they impacted upon the willingness of
users to share information on SNS. In order to achieve this main research objective,
Structural Equation Modeling was considered to be suitable. The generated model is
expected to be a model that is both substantively meaningful and statistically well-
fitting (Joreskog & Sorbom 1996).
Structural Equation Modeling (SEM) is a multivariate technique combining aspects
of multiple regression (examining dependence relationships) and factor analysis
(representing unmeasured concepts-factors with multiple variables) to estimate a
series of interrelated dependence relationships simultaneously (Hair Jr et al. 2006;
Schumacker & Lomax 2004). SEM also integrates other techniques such as recursive
path analysis, non-recursive econometric modeling, ANOVA, analysis of covariance,
principal component analysis and classical test theory (Holmes-Smith 2000). In
addition, SEM is also known as path analysis with latent variables and is now a
regularly-used method for representing dependency (arguably ‗causal‘) relations in
72
multivariate data in behavioural and social sciences (Kripanont 2007; McDonald &
Ho 2002).
A structural equation model or path model depicts the structural relationships among
constructs (Sharma 1995). In other words, it is a model of relationships among
variables (Hayduk 1987), and is a statistical methodology that takes a confirmatory
(i.e. hypothesis-testing) approach to the analysis of a structural theory relating to
some phenomenon with two important aspects: (1) the causal processes under study
are represented by a series of structural equations, and (2) these structural relations
can be modeled pictorially to enable a clearer conceptualisation of the theory under
study (Byrne 2001). When compared to other multivariate techniques, it has four
significant benefits over those techniques (Byrne 2001):
1) SEM takes a confirmatory approach rather than an exploratory approach to
the data analysis, although SEM can also address the latter approach. SEM
lends itself well to the analysis of data for the purposes of inferential
statistics. On the other hand, most other multivariate techniques are
essentially descriptive by nature (e.g. exploratory factor analysis) so that
hypothesis testing is possible, but is rather difficult to do.
2) SEM can provide explicit estimates of error variance parameters, but
traditional multivariate techniques are not capable of either assessing or
correcting for measurement error.
3) Data analysis using SEM procedures can incorporate both unobserved (i.e.
latent) and observed variables, but the former data analysis methods are based
on observed measurements only.
73
4) SEM methodology has many important features including modeling
multivariate relations for estimating point and/or interval indirect effects,
although there are no widely and easily applied alternative methods for these
kinds of features.
In particular, SPSS version 19.0 was used to input and conduct preliminary analyses
of data (see Chapter 5) together with an SEM software package called AMOS
version 19.0. This was used to test the model fit. Structural Equation modeling
techniques demonstrate and test the theory of representation with a model that shows
how measured variables combine together to represent construct validity (Hair Jr et
al. 2006). Furthermore, Confirmatory Factor Analysis (CFA) enables the researcher
to illustrate how well the measured variables are represented in the constructs, and it
also shows how the results are combined with construct validity tests in order to
maintain a good understanding of quality of the measurement according to Hair Jr et
al (2006).
4.10 Conclusion
This chapter outlined the methodology used for the research design and data
collection within this study. It described and justified how the data was prepared for
analysis and was then analysed within each hypothesis. The author described the
process of ensuring validity and reliability in order to come up with the final
instrument. Wherever possible, existing measures that were proven to be reliable and
valid were adapted from prior studies. However, rather than setting out to validate
the measures that have already been validated many times in various IT adoption and
online electronic business studies, the purpose of this study has been to develop or
modify a new set of measures in the SNS context, provided there is support from the
74
existing literature. As mentioned in the data analysis strategy above, Chapter Five
will discuss the data analysis procedure using SPSS 19.0 and AMOS 19.0 and
present the results.
75
Chapter V: Results and Analysis
5.1 Introduction
This chapter contains the empirical results of this study and divides the data analysis
measurements into the following steps. The first step is the descriptive analysis to
describe the respondents‘ characteristics, and the last part contains the statistical data
analysis using SPSS, followed by a presentation of the results. The results of the
SEM analysis are presented to test the interrelationships among SNS users‘
experiences on how they are influenced by knowledge, privacy concerns and trust
and how it affects their willingness to share digital identities.
To investigate, the following objective of this study was established to determine the
relationship among the construct and shows effect on willingness to share digital
identities in SNS:
To what extent do privacy concerns and trust influence user willingness to
share digital identities on social networking sites?
5.2 Data Quality and Characteristics of Respondents
Data quality and its suitability for analysis were ensured through careful inspection
and review of the data. The attributes of respondents consisted of six major variables
including gender, marital status, age, education, occupation, number of friends on
SNS, the number of SNS accounts; and the results of these questions and answers
will be discussed in detail with reference to each variable in the below Figure. The
automatic survey system showed that 196 respondents started filling in the
questionnaire, however, only 155 were completed sufficiently to allow for adequate
76
future research.
This research follows Hair Jr et al‘s (2006) approach to missing data patterns and,
according to their guidelines, adherence to statistical assumptions, identification of
outliers and a review of skewness and kurtosis were inspected. The data was
carefully reviewed and tested, and the results showed that the data was suitable for
further analysis. Two types of missing data patterns were examined: missing data for
each case and missing data for each variable. While several missing data of both
types were found, the examination revealed that missing data was not a problem.
Among the 155 respondents 54.2% were male, and from this it seems that almost
equal numbers of users of each gender user were accessing the SNS, as shown in
Table 5-1. This was a surprising result, because most of the online groups to which
the researcher promoted the survey had female moderators. The majority of the
respondents were aged between 18–30 years. A graphical representation of the age
distribution of the respondents is shown in Figure 5.1. A total of 81.95% of
participants were between the ages of 18 and 40, while 16% were more than 40 years
of age. However 1.29% of users did not disclose their age.
77
Figure 5.1 Respondent age groups percentiles
The length of time that participants had been using the Internet is shown in Figure
5.2. More than 36% of the participants had more than 5 years of Internet experience
(n=155). Only 1.6% of respondents had used the Internet for two years or for less
than two years. This result shows that the number of Internet users is increasing year
by year.
78
Figure 5.2 Use of Internet
Figure 5.3 describes how long the participants had been using SNS. Altogether, more
than 30% of the respondents had been members of and using SNS for the last 2 to 4
years. More than 25% had been members for the last 5 to 6 years. More than 7% of
respondents had used SNS for more than 8 years. However less than 3% of users did
not want to disclose how long they had been using SNS.
79
Figure 5. 3 Use of social networking sites
All of the participants belonged to at least one SNS. The number of SNS accounts
opened by respondents at the time of their participation in the survey is graphically
displayed in Figure 5.4. More than 31% of respondents had two different SNS
accounts. More than 29% of respondents have one SNS account and less than 12% of
respondents have more than four SNS accounts. A few respondents did not want to
disclose how many SNS accounts they had registered.
80
Figure 5.4 Current number of accounts on different SNS
The graphical representation in Figure 5.5 shows how often users logged into their
SNS accounts. While 46.5% of respondents said they had visited the site several
times a day, 23.2% visited only once per day. A total of 15.5% of respondents did
not visit the site each day, but they did visit the site within a week, while a further
3.2% of users logged onto the SNS once a month. Fewer than 2% of respondents did
not want to disclose how often they visited SNS.
81
Figure 5.5 Number of visits to social networking sites
SNS are growing exponentially, and the majority of respondents became aware of
SNS through friends, news and media. Most of the respondents (61.9%) responded
that they became aware about the SNS for first time from their friends, as shown in
Figure 5.6. Due to the frequent use of Internet, 39% of participants gained awareness
of SNS through their use of the Internet and 9.7% of users came to know about SNS
through news and the media. However, less than 1% of users did not want to disclose
how they came to know about SNS.
82
Figure 5.6 Source of original knowledge about SNS
Graphical representation in Figure 5.7 shows that each of the respondents had a
number of friends in their network; 27.1% of participants had more than 150 friends
in their network in their SNS account; and 23.9% of respondents reported more than
200 friends listed in their SNS account. Of the respondents, 16 % reported they had
more than 100 friends listed; 12.9% of users reported that they had less than 100
friend connections; and 9% of users had more than 300 friends listed in their SNS
accounts. Less than 2% of all respondents did not want to disclose how many friends
they have in their SNS account.
83
Figure 5.7 Average numbers of friends in each social networking account
Users cited different reasons for visiting SNS, as displayed in Figure 5.8. Most of the
respondents (56.1%) used SNS to keep in touch with friends and family; followed by
21.9% of participants who were looking for new friends on SNS. More than 11.6%
of respondents used SNS for the purposes of professional or work related tasks. More
than 10% of participants used SNS for professional reasons, to socialise or to access
the latest news.
84
Figure 5.8 Purpose for visiting social networking sites
Figure 5.9 displays graphically the varying levels of education of the questionnaire
respondents. From the results, it is clear that 7.7% of respondents had completed a
Doctoral degree, 31% reported they had completed a Master‘s degree; 30.3% had
completed a Bachelor‘s degree; more than 25% reported they had completed a
Diploma; and less than 6% reported their highest level of education as the
completion of high school.
85
Figure 5.9 Education status of respondents
Other attributes of the respondents that were analysed included their gender, marital
status and occupation, as shown in Table 5-1. Of the 155 respondents, 56.1% were
married; 40% of participants were single; and 1.9% of participants mentioned that
they had a partner or were engaged to be married. However, 1.9% of users did not
want to disclose their marital status. The results also analysed the different
occupations of the respondents, and the findings were that the majority of
participants were students (47.7%), followed by professionals 24.5%, while 12.9% of
respondents were from an academic background.
86
Table 5-1 Characteristics of the respondents
Personal Information Frequency Percentage
Gender
Male 84 54.2%
Female 71 45.8%
Marital status
Single 62 40%
Married 87 56.1%
Others 3 1.9%
Do not want to disclose 3 1.9%
Occupation
Student 74 47.7%
Academic 20 12.9%
Manufacturing/construction 1 0.6%
Profession 38 24.5%
Business 6 3.9%
Self-employed 5 3.2%
Retiree 2 1.3%
Others 6 3.9%
Do not want to disclose 3 1.9%
5.3 Confirmatory Factor Analysis (CFA)
The confirmatory factor analysis (CFA) was adapted to verify the adequacy of the
item to factor associations and the number of dimensions underlying the construct
(Thompson & Daniel 1996). CFA is a way of testing how well measured variables
represent a smaller number of constructs (Thompson 2004). One of the biggest
advantages of CFA is its ability to assess the construct‘s validity (Hair Jr et al. 2006).
87
Validity is defined as the extent to which the research was accurate.
The constructs of SNS experience, privacy concerns, trust and willingness to share
digital identities were adopted from previous studies as stated factors; and to further
validate the constructs, confirmatory factor analysis was used within structural
equation modelling. Confirmatory factor analysis in structural equation modelling
gives a more accurate depiction of the relationship between the dimensions since
measurement error is taken into consideration (Hair Jr, Anderson, Tatham & Black
1998). The validity and reliability of this research was found to be significant, as
shown in Table 5-2. However, this may not be true if the construct and paths are put
together in an overall SEM framework. Therefore, a more rigorous method of
statistical analysis is used to show the interactions between dependent and
independent variables.
Construct validity is made up of three important components, namely, factor
loadings, variance extracted, and construct reliability. The standardised loading
estimates should be 0.5 or higher and, ideally, 0.7 or higher. With CFA, the average
percentage of variance extracted (VE) among a set of construct items can be
calculated simply using standardised loadings that are squared before summing them
and dividing by the total number of items (N). The measurement of Construct
Reliability (CR) is quite similar to the VE where CR is computed from the squared
sum of factor loadings and a sum of the error variance for each construct; whereas
error variance is calculated by taking one minus factor loading square: the
mathematical calculation is shown below (Hair Jr et al. 2006, p. 612).
88
Error Variance = 1-
CR =
Where λi is standardised loadings obtained from each latent construct, N is number
of item and is the measurement of error for each indicator. The measurement error
is 1.0 minus the reliability of the indicator, which is the square of the indicator‘s
standardised loading (Hair Jr et al. 2006).
This is in accordance with a suggestion by Hair Jr et al (2006) that the criteria of
construct validity are as follows:
i) Standardised loading estimates ≥ 0.5
ii) Variance extracted (VE) ≥ 0.5
iii) Construct reliability (CR) ≥ 0.7
On the other hand, to access overall reliability the Cronbach‘s alpha coefficient for
each dimension and construct was calculated as shown Chapter IV Table 5-2. An
alpha score that is greater than 0.70 was considered to be acceptable (Nunnally 1994;
Sekaran 2003). The alpha value for each item for this model ranged from 0.816 to
0.906, and all scales had construct reliabilities above 0.7, therefore, data collection
for each construct are reliable and have validity for further analysis.
89
Table 5-2 Reliability statistics
Details of the measurement for each construct item are separately tested by the CR,
VE and alpha value in order to check the reliability and validity of data (see
Appendix E). Item values of less than 0.7 are considered as unreliable for study so
these items are deleted and the items with higher than 0.7 (listed in Table 5-3) are
considered for further data analysis.
The results of CFA for all variables are shown in Table 5-2. Construct privacy has
one factor with five items (PC1, PC2, PC3, PC4, PC5) with factor loadings of .61,
.67, .77, .75 and .63 respectively as shown in Figure 5-10. CR for privacy is greater
than 0.7 so values are adequate for this study.
The construct trust has also one factor with 11 items (T1 to T11) where most of the
items have a factor loading of more than 0.7, although some items had a factor
loading of less than 0.5 and these items have been deleted and left for future
research. VE and CR for the trust construct were higher than 0.5 and 0.7
respectively.
Construct Cronbach's Alpha Cronbach's Alpha Based on Standardised
Items
Number of
Items
SNS Experience .753 .757 4
Privacy Concerns .816 .817 5
Trust .881 .882 11
Willingness to share digital
identity .906 .911 13
90
The construct ‗willingness to share digital identities‘ has one factor that consists of
thirteen items (DI1 to DI13) for which the factor loading is shown in Table 5-2. VE
and CR for this factor are higher than 0.5 and 0.7 respectively.
The construct ‗SNS experience‘ consists of one factor with four items (SK1 to SK4)
with factor loadings .76, .71, .66, .70 respectively. CR is greater than 0.7. Details of
the statistical table SPSS output are attached in Appendix E.
91
Table 5-3 Convergent validity of the model variables
Factor Items Standardised
Factor Loadings
Corrected
Item Total
Correlation VE
Cronbach‘s
Alpha If
Item Deleted CR
Privacy
concerns
0.816
PC1 0.61 0.564
0.480
0.793
PC2 0.66 0.612 0.779
PC3 0.78 0.652 0.767
PC4 0.76 0.636 0.771
PC5 0.61 0.573 0.791
Trust
T1 0.66 0.620
0.614
0.884
0.886
T2 0.81 0.769 0.846
T3 0.84 0.762 0.848
T4 0.85 0.789 0.843
T6 0.74 0.678 0.868
SNS
experience
SNS1 0.76 0.732
0.510
0.710
0.707
SNS2 0.71 0.690 0.705
SNS3 0.66 0.648 0.701
SNS4 0.70 0.685 0.707
Willingness to
share digital
identity
DI2 0.50 0.468
0.567
0.892
.869
DI7 0.70 0.633 0.852
DI8 0.85 0.777 0.827
DI9 0.91 0.823 0.821
DI10 0.79 0.737 0.833
DI13 0.71 0.663 0.850
92
Figure 5.10 Measurement fit model (Source: The Author)
5.4 Dimensional Level Analysis – The Measurement Model
To further validate the constructs, confirmatory factor analysis was used within
structural equation modelling. Confirmatory factor analysis in structural equation
modelling gives a more true relationship of the dimensions since the measurement
error is taken into consideration (Hair Jr et al. 2006).
According to Anderson and Gerbing (1988), the measurement model specifies how
the latent variables or hypothetical constructs are measured in terms of the observed
93
variables, taking into account specification error. Before starting on testing the
proposed measurement models for each construct, various fit indices need to
discussed. Although there are number of fit indices, Maruyama (1998) argued that
there is no single test that best describes the fit of a model. He categorises fit
measurement as corresponding with three types: absolute, relative and adjusted
indexes.
Absolute Fit indexes provide information about how closely the model fits compared
to perfect fit (Maruyama 1998). This can be measured mainly by (chi-square)
test. A low value, which would have a p-value greater than 0.05, indicates that
the actual and predicted values are not significantly different. Another index, relative
fit index, also known as Comparative Fit Index (CFI), is a measure of how the model
compares with other possible models with the same data (Maruyama 1998). CFI
provides an estimate of the model‘s relative misfit to a baseline model. Higher
numbers indicate a lower misfit. Normed Fit Index (NFI) also compares the
theoretical model to a baseline model. A recommended value of fit for both NFI and
CFI is 0.90 and a model will be marginally fit at greater than 0.8 (Hair Jr et al. 2006).
Another index is Goodness of Fit Index (GFI) that tells that proportion of the
variance in the sample variance-covariance matrix is accounted for by the model.
This should exceed 0.9 for a good model and 0.8 for marginal fit. For the full model,
a perfect 1 would be recommended. AGFI (Adjusted GFI) is an alternate GFI index
in which the value of the index is adjusted for the number of parameters in the
model. According to Etezadi-Amoli and Farhoomad (1996) for a good model AGFI
should be greater than 0.8 and near to GFI.
94
Another fit statistic is the Root Mean Square Error of Approximation (RMSEA), as a
measure of fit. Joreskog and Sorborn (1996) suggests that a value of the RSMEA of
about 0.06 or less would indicate a close fit of the model in relation to the degrees of
freedom although this figure is based on subjective judgement and cannot be
regarded as infallible (Arbuckle 2006, p. 496; Shin 2010a).
The following Table 5-4 shows the details of the Confirmatory Factor Analysis in the
structural equation model. To determine the best-fitting model, this study used the
chi-square difference tests as well as the Comparative Fit Index. Bagozzi and
Edwards (1998) have made the suggestion that the CFI was particularly useful for a
small sample size because it, unlike the chi-square statics, operates independently of
sample size. Table 5-4 represents the result of Goodness of Fit Indexes of the
Measurement Model. The Model Fit Index values prove the model is fit where the
chi-square/d.f (χ2/df) is less than 5, GFI and CFI have met the acceptance criteria.
Table 5-4 Goodness of Fit Indexes of the Measurement Model
Fit Indexes Criteria Indicators Acceptability
Chi-Square (χ2) >0.05 238.378(.000) Acceptable
Chi-Square/d.f (χ2/df) <5.00 1.454 Acceptable
Goodness of Fit Index
(GFI) >0.90 .867 Marginal
Adjusted GFI >0.80 .829 Acceptable
Comparative Fit
Index (CFI) >0.90 .939 Acceptable
Normed Fit Index
(NFI) >0.90 .831 Marginal
95
Fit Indexes Criteria Indicators Acceptability
Incremental Fit Index
(IFI) >0.90 .941 Acceptable
Tucker-Lewis
coefficient (TLI) >0.90 .930 Acceptable
Root Mean Square
Error of
Approximation
(RMSEA)
<0.06 .054 Acceptable
5.5 Descriptive Statistics and Correlation for All Variables
Table 5-5 shows the means, standard deviations and correlation matrix for all
variables of the research models that contain privacy concerns, trust, SNS experience
and willingness to share digital identities.
Based on the data shown in Table 5-5, it is suggested that for the construct of privacy
concerns, respondents tend to perceive a relatively higher degree of agreement where
the mean value for privacy concerns is 5.16 in a 7-point Likert scale. These results
indicate that the respondents have high levels of privacy concerns. In addition, for
the construct of trust in SNS, respondents have neutral level of agreement on the
measurement factors with a mean score 4.52. This indicates that users have relatively
high levels trust of with SNS and with its users. The results of SNS experience
scored over 3, which indicates that respondents were using SNS for over 5 years and
using SNS at least once a day. The construct of willingness to share digital identity
had a mean score of 4.88 which indicates that in 7 point scale, the respondents had a
low level of intention to share all aspects of their digital identities in SNS.
96
Table 5-5 Descriptive Analysis and Correlation
Correlation
Mean Std Dev Privacy Trust Identity SNS
Experience
Privacy 5.16 1.083 1
Trust 4.52 .955 -.035 1
Identity 4.88 1.136 -.336** .360** 1
SNS
Experience 3.82 .893 -.038 -.089 -.033 1
** Correlation is significant at the 0.05 level (2-tailed).
5.6 Structure Equation Modelling (SEM)
The major purpose of using SEM is to test and estimate the relationships between
research constructs and to provide estimates of the strength of all hypothesised
relationships between variables in the theoretical model (Anderson et al. 1988).
A Structural Equation Model (SEM) can provide information about hypothesised
impact, both directly from one variable to another, and also indirectly through other
variables (Maruyama 1998). Additionally, Baumgartner and Homburg (1996, p. 141)
argued that using structural equation models can be specified to investigate
measurement issues to examine structural relationships among sets of variables or to
accomplish both purposes simultaneously. The use of SEM for this research is to
confirm the hypothesised paths and overall fit of the theoretical model.
There are two distinct parts of SEM: measurement model and structural model
(Maruyama 1998). The structural model defines relationships between the
unobserved variables. The constructs or unobserved variables for this study have
been statistically validated through measurement model as shown in previous section
(Fig 5-10), and as such the model that will be used in this section is the structural
model. To analyse this SEM, we used SPSS AMOS 19 software for data analysis
purpose and details of the output of AMOS is shown in Appendix - E.
97
Importantly there are several Goodness of Fit measures that can be used to assess the
outcome of SEM analysis, they include the Root Mean Square Error of
Approximation (RMSEA) which is based on chi-square values and measure the
discrepancy between observed and predicted values per degree of freedom (a good
model has an RMSEA value of less than 0.06) (Joreskog et al. 1996; Shin 2010a); the
Comparative Fit Index (CFI), which compares proposed model with baseline model
with no restrictions (a good model should exhibit a value greater than 0.90) and
Goodness of Fit measures, which compare the sample and model implied variance
covariance matrices such as the standardised root mean square residual (SRMR) and
a value less than 0.08 (Bentler 1990; Shin 2010a) is considered as a good fit. The
adjusted goodness of fit index (AGFI) considered the greater the value better.
Moreover, the main criteria for SEM are the following:
Chi-square value should be higher than 0.05
Chi-square/ degree of freedom should be smaller than 2
Goodness of Fit Index should be higher than 0.80
Root Mean Square of Standardised Residual should be smaller than 0.060
Furthermore, based on the standardised structure equation model (Figure 5.11)
overall Goodness of Fit Indices and path coefficient are shown in Table 5-6 and
Table 5-7 respectively.
98
Figure 5.11 Standardised Structure equation model path diagram (Source: The
Author)
99
Figure 5.12 Final Structural equation model of impact of privacy concerns and
trust in SNS for willingness to share digital identities (Source: The Author)
100
Table 5-6 Goodness-of-Fit indices of structural model
Fit Indexes Criteria Indicators Acceptability
Chi-Square (χ2) >0.05 243.770(.000) Acceptable
Chi-Square/d.f (χ2/df) <5.00 1.477 Acceptable
Goodness of Fit Index
(GFI) >0.90 .865
Marginal
Adjusted GFI >0.80 .828 Acceptable
Comparative Fit Index
(CFI) >0.90 .936
Acceptable
Normed Fit Index
(NFI) >0.90 .828
Marginal
Incremental Fit Index
(IFI) >0.9 .937 Acceptable
Tucker-Lewis
coefficient (TLI) >0.9 .926 Acceptable
Root Mean Square
Error of Approximation
(RMSEA)
<0.06 .056 Acceptable
5.6.1 Overall Model Fit
According to Bentler (1990; 2007, p. 827) and Jorsekog and Sorbom (1996) GFI,
AGFI, RMSEA, IFI and TLI are useful for describing the best fit model. The result
represented in Table 5-6 shows that the Chi-square value of this research model is
243.770 with degree of freedom 165 significant at <0.001. Table 5-6 shows the
details of goodness of fit indices for structural model where Chi-square/ d.f is 1.477
that is less than 5. GFI value is .865, according to Hair Jr et al (2006) greater than 0.8
is marginal acceptance, to become best fit GFI should be >0.90 and AGFI >0.80
(Etezadi-Amoli et al. 1996; Shin 2010a). The model also supports the CFI is 0.936
and NFI is 0.828 which must be greater than 0.9 for a best fit and greater than 0.8 for
101
marginal acceptance according to Hair Jr et al (2006). The model has also get a better
RMSEA value as shown in Table 5-6 which is 0.053 and value of RMSEA should be
less than 0.06 according to Joreskog & Sorborn (1996); and Bentler (1990)
respectively. The incremental fit index (IFI) and Tucker-Lewis index (TLI) has
values greater than 0.9, which indicates higher levels of Goodness of Fit. Overall the
results of Goodness of Fit indicate that the model was a good fit.
5.6.2 Path Results
Table 5-7 Path Coefficients for structural model
Estimate S.E Critical
Ratio p
SNS experience Willingness to
share digital identities -0.257 1.101 -1.021 .307
Privacy concerns Willingness to
share digital identities -0.255* .109 -2.474 .013
Trust Willingness to share digital
identities 0.282** .089 2.962 .003
Privacy concerns Trust -0.058 .166 -2.108 .536
Note: * p< 0.05, **p<0.003
To test the research hypothesis, path analysis is performed using AMOS 19 software
tools and based on the SEM of this study there are four paths and the details of the
path analysis and strength of the relationship between constructs are shown in Table
5-7 and based on this results, the summary of testing of the hypothesis is shown in
Table 5-8. This study has established four hypotheses: dependent relationships were
established between the constructs of privacy concerns, trust, SNS experience and
willingness to share digital identities and out of the four, two hypotheses were found
to be supported.
102
According to Hair Jr et al (2006), to support the hypothesis each path should be
significant at level of 0.005 and its C.R. must be greater than 1.96. No significant
relationship was found between SNS experience and willingness to share digital
identities, so there is a weak relationship and, therefore, hypothesis (H1) was not
supported for this study—which would suggest that the SNS experience does not
influence people‘s willingness to share digital identities on SNS. Also, there is no
significant difference between privacy concerns and trust; moreover, there is very
low standardised estimated path coefficient for the relationship between privacy and
trust. So, this finding strongly rejects the hypothesised relationship between privacy
concerns and trust. However, some previous studies have shown that there must be a
relationship between them in order for them to share information. This will be
discussed in detail in the next chapter. All of the paths between other variables have
shown that they are strongly significant and C.R. are higher than 1.96 which means
that all the hypotheses implemented under this path are supported by this study.
Table 5-8 Results of hypotheses testing
Hypothesis Results of study
H1 Users with high levels of SNS Experience will be more
willing to share digital identities Not supported
H2 Users with higher level of privacy concern has lower
level of trust in SNS. Not Supported
H3 Users with higher level of trust will be more willing to
share digital identities in SNS. Supported
H4 Users with higher levels of privacy concerns will be
less willing to share digital identities. Supported
5.7 Conclusion
A great deal of analysis has been undertaken in this chapter of the responses to the
questionnaire questions. These results have been verified and tested, and then
presented in this chapter. The analysis has focused on reliability analysis, factor
103
analysis and structural equation model of this study. To test reliability, the
Cronbach‘s alpha test was run for which the results need to be greater than 0.7 and
the findings show that most of the items are reliable for this study. The researcher
also calculated variance, for which the data extracted must be greater than 0.5 and
these findings were significant for this study. Confirmatory Factor Analysis for each
variable with a loading factor greater than 0.50 were used for the analysis, and those
with a result below 0.50 loading factor were deleted. The different types of Goodness
of Fit indices were conducted to measure the fitness of model such as GFI, AGFI,
CFI, NFI, IFI, TLI and RMSEA.
The value of AGFI, CFI, IFI, TLI and RMSEA are respectively .828, .936, .937, .926
and .056, which indicates that the model is a better fit. The GFI and NFI has a value
.865 and .828 respectively, which are at the marginal acceptance level.
The outcome of measurement model is acceptable and the main purpose of SEM was
the testing of the proposed hypothesis stated in Chapter III, and hypotheses were
tested using path analysis. Estimation values show that the extent of SNS experience
has a negative impact upon users‘ willingness to share digital identities. Similarly,
privacy concerns have negative effects on trust and willingness to share digital
identities. However, trust is shown to have a positive effect on willingness to share
digital identities. The results show that two hypotheses are fully supported, while two
hypotheses could not be supported.
All of the analysis was aimed at understanding the impact of privacy and trust on
users of SNS in sharing digital identities. The following chapter provides a
104
discussion of the implications of accepting and rejecting hypotheses in terms of the
findings of the path analysis, as well as an examination of the limitations of the study
and suggestions for areas of further research.
105
Chapter VI: Conclusion
6.1 Introduction
This chapter will discuss the findings associated with the statistical analysis of the
hypothesised relationships of the research model for this study. Each of the
hypotheses developed in Chapter III will be reviewed, followed by a subsequent
discussion of the findings. It should be noted that the Kline (2010) method has been
used to discuss the impact of privacy concerns and trust on using SNS. That is, if the
absolute value of the standardised path coefficient in an structural equation model is
less than .10, this will indicate a ‗small‘ impact, while a coefficient of around .30 will
indicate a ‗medium‘ impact, and a coefficient greater than .50 will indicate a ‗large‘
impact.
This study shows that there is an impact of privacy concerns, trust and SNS
experience upon the willingness of SNS users to share information. However, there
is a positive and negative effect between different variables.
This study carried out analyses of four different types of relationships, namely,
privacy concerns, trust, SNS experience and willingness to share digital identities in
SNS. The literature confirms that users with high levels of Internet experience have
more confidence, and are willing to share more information; and for those with low
levels of experience, the opposite was found to be true. The following sections
discuss these findings in more detail, including the possible limitations of the study
and suggestions for future research.
106
6.2 Summary of the study
This section provides a summary of the research problem and general research
question investigated in this study, the research hypotheses which were tested, and
the research method used in this study. The key findings of descriptive demographic
data analyses and hypotheses testing are then summarized.
6.2.1 Research problem
Dwyer, Hiltz and Passerini (2007) found that privacy concerns and trust were major
issues for users of SNS. As the world has become more digitised, protection of the
privacy of the user has become more complicated. Nysveen and Pedersen (2004)
found that privacy concerns in SNS were similar to those found for users of
electronic commerce in terms of concerns by users about providing personal
information. Pavlou and Fygenson (2006) established that customers who were
using SNS and online business transactions were worried about their personal data
regarding its security and privacy, however, their research suggests that customers
conducting online business transaction are more concerned about their privacy in this
domain than in sharing the same personal information on SNS. Furthermore, they
suggest that despite their concerns about security, users remained willing to share
their personal information within SNS. However, these researchers were unable to
justify the reasons why users were more willing to share their personal information
within SNS.
Previous studies conducted by Lo (2010) on privacy concerns in SNS suggest that
knowledge and experience may affect the perception of users about privacy issues as
they relate to the sharing of personal information. However, his study did not justify
107
whether users‘ SNS knowledge and experience play critical role to influence users‘
privacy concerns and trust to share personal information in SNS.
Moreover, very few previous studies have investigated privacy concerns and trust in
e-commerce (e.g. Lio 2005)—which is almost a similar concept to SNS—or on how
the users‘ privacy concerns and trust impact on sharing users‘ identities on SNS.
Based on the research problem the main research question of this study is:
To what extent do privacy concerns and trust influence users‘ willingness to share
information about their digital identity within Social Networking Sites?
Therefore, the main research objectives that underpin the general research question
of this study are:
To examine the impact of users' experience with SNS upon their
willingness to share their digital identities.
To examine the influences of privacy concerns as they relate to the
trust needed for users to share their digital identities within SNS.
To examine the influence of trust about SNS upon users‘ willingness to
share their digital identities.
To examine the impact of privacy concerns upon the willingness of
users of SNS to share their digital identities.
6.2.2 Research hypotheses
The four hypotheses were formulated from the four research questions above after
being justified and grounded in the existing relevant literature on SNS experiences
108
and knowledge, privacy concerns, trust and willingness to share digital identities.
The four hypotheses are as follows:
H1: Users with high levels of SNS experience will be more willing to share digital
identities.
H2: Users with higher levels of privacy concern have lower levels of trust in SNS.
H3: Users with higher levels of trust will be more willing to share digital identities
on SNS.
H4: Users with higher levels of privacy concerns will be less willing to share digital
identities.
These four hypotheses test how the users‘ SNS experience and knowledge, privacy
concerns and trust will impact on sharing the users‘ digital identities in SNS.
6.2.3 Research Methodology
This is explanatory research using a quantitative approach to test the research model
which examines the relationship between the independent variables such as SNS
experiences and knowledge, privacy concerns and trust; and the dependent variable
willingness to share digital identities. For the purpose of this study, data is collected
from Facebook Forum users where the numbers of registered users are publicly
available in the forum and the total numbers of participants in the online survey were
169 users. Facebook Forum is a community forum where users are able to share,
discuss and comment on anything related to Facebook or other SNS such as a new
SNS feature, its advantages and disadvantages, an application, games, new privacy
settings, scams, photos, videos, events, and numerous other general topics where
users can request help. There are also forum contests, general chat areas and an
109
interactive Chatbox, where online members gather to chat together in real-time
(Forum 2011). Facebook Forum was selected for this study because it is a subset of
Facebook and it will improve the credibility of the results because participation
includes users from different backgrounds including application developers,
professionals, students and social workers, which can represent a much broader
population of SNS users.
6.2.4 Conclusions about descriptive demographic data
Of the 155 respondents to the survey, 54.2% were male and 45.8% were female. A
total of 81.95% of participants were aged between 18-40, while 16% were more than
40 years of age. From this background of age, it can be seen that most of the
participants were young; and 47.4% participants were students—with the remainder
involved in various professions or other categories (see Table 5-1). A total of 96% of
participants were concerned about their privacy with 76.1% of participants
modifying their privacy settings; however, only 35.5% had read about privacy
policies or terms and conditions of social networking sites.
6.2.5 Conclusions about SEM model fit
The different types of Goodness of Fit indices were conducted to measure the fitness
of model such as GFI, AGFI, CFI, NFI, IFI, TLI and RMSEA. The value of AGFI,
CFI, IFI, TLI and RMSEA are respectively .828, .936, .937, .926 and .056, which
indicates that the model is a better fit. The GFI and NFI has a value .865 and .828
respectively, which are at the marginal acceptance level. The outcome of
measurement model is fit and the main purpose of SEM was the testing of the
proposed hypothesis stated above, and hypotheses were tested using path analysis.
110
6.2.6 Conclusions concerning Results of Research Hypotheses Tests
H1: Users with high levels of SNS experience will be more willing to share digital
identities.
The validity and reliability of four dimensions of the construct SNS experience were
found to be significant, as shown in Chapter V. However, this did not prove to be
true when the construct and paths were put together in an overall comprehensive
framework as shown in Figure 5-11 (Chapter V). The ‗p‘ value between SNS
experience and willingness to share digital identities was not found to be significant.
As such, the structural equation model did not support this hypothesis. However, the
results of path estimations show that there is negative relationship between individual
users‘ SNS experiences and the sharing of their digital identities.
A plausible explanation for this result could be found when it is considered that the
majority of SNS users were university students and were aged under 30 years
(54%)—which shows that this is the age at which users were influenced by friends
(more than 60%) and were susceptible to peer pressure.
The rejection of this hypothesis is, however, consistent with the problems inherent in
the successful measurement of SNS experiences, which has not kept pace with
theoretical developments. Novaka, Hoffman and Yung (2000) argued that Internet
experiences were usually defined as general experience with Web sites and not as
experiences with one particular Web site. Prior experience has been found to be an
important determinant of behaviour. Long term experience effects and changes the
behaviour of the user according to Ajen and Fishbein (1980). The statistical analysis
of the study shows that the majority of respondents had just started using SNS during
111
the last four years, which indicated that respondents had less SNS experience.
Moreover, the negative value of estimated path coefficient shows that the SNS
experience has a negative effect on willingness to share digital identities.
H2: User with higher level of privacy concern has lower level of trust in SNS.
The validity and reliability dimensions of the construct were found to be significant.
The path analysis result shows that there is no significant (p <0.005) effect between
privacy concerns and trust, but the C.R. is greater than 1.96, which shows the
hypothesis is partially supported by the study results. However, there is no
significant result found between privacy concerns and trust; and, additionally, there
is very low standardised estimated path coefficient for the relationship between
privacy and trust. Thus, this finding strongly rejects the hypothesised relationship
between privacy concerns and trust. The negative value estimations on path analysis
show that if the user has high privacy concerns, then they have low trust and vice-
versa, which is a logical outcome. Protections of privacy, as well as the mechanism
to protect trust, are critical to designers and developers of SNS systems, according to
Joinson, Reips, Buchanan and Schofield (2010). According to Grewal, Munger, Iyer
and Levy (2003), privacy policy statements appear to be most beneficial to the web
developer who is seeking to increases users‘ trust (Meinert, Peterson, Criswell &
Crossland 2006b). However, this study‘s findings are inconsistent with previous
findings by Liu et al (2004) which found that privacy concerns had positive impacts
on trust in electronic commerce.
112
H3: Users with higher levels of trust will be more willing to share digital identities
on SNS.
There is a strong significant relationship between trust and willingness to share
digital identities on SNS. As a result, the reliability and validity dimensions of the
construct were found to be significant, and also the path estimations between two
constructs were found to be significant at p < 0.003 and C.R. is greater than 1.96. As
such, the structural equation model strongly supports this hypothesis.
This statistical result suggests that high levels of trust in SNS by users had the effect
of heightening the probability that users would be willing to share digital identities
on SNS. The business market research conducted by Grayson and Ambler (1999)
found that trust has a central role in building long–term relationships with their
clients. Also, some of the findings regarding trust and identity sharing in this study
are consistent with prior findings. For example, studies by Lo et al (2010), Dwyer at
al (2007) and Fogel and Nehmad (2009) found that trust has a positive impact on
users. Thus, this finding suggests that users will share more information on SNS if
they have a high level of trust in SNS.
H4: Users with higher levels of privacy concerns will be less willing to share digital
identities.
Concerns about privacy have a strong impact upon users‘ willingness to share
identities, and this impact is statistically significant, as shown in Figure 5-11. The
path analysis shows that there is a significant path (p<0.05) and C.R. is greater than
1.96, which indicates that the structural equation model strongly support this
hypothesis for this study. Also, reliability and validity dimensions of the construct
113
were found to be significant, which makes the structural equation model fit for this
study. Additionally, the strong negative impact of estimation path results show that
high levels of privacy concerns make users less inclined to share digital identities on
SNS, and low levels of privacy concerns create a high level of willingness to share
digital identities on SNS. However, the findings of this study are inconsistent with a
prior finding about privacy concerns by Metzger (2004), but the finding is consistent
with Lo et al (2010). However, unlike Lo et al (2010), this study has found that there
is a significant relationship between privacy concerns and users‘ willingness to share
digital identities.
6.3 Contribution of Study
This research study makes several practitioner contributions, especially for the
literature, SNS developers and users. The researcher summarises the contributions
and implications to practitioners as follows:
6.3.1 Contributions to the Literature
This study provides at least two contributions to the literature. Firstly, it provides a
better understanding and new insights into privacy concerns and trust, and their
effect on users‘ willingness to share their digital identities on SNS. Secondly, this
study will enrich existing literature regarding user experiences and knowledge of the
Internet, especially in relation to SNS‘s effect on users‘ willingness to share identity
information. In addition, this study also shows the relationship between trust and
privacy.
114
Furthermore, from the analysis of results, it can be seen that the extent of users‘ SNS
experience and knowledge has a negative effect upon their willingness to share
digital identities. Similarly, users with higher levels of privacy concerns were less
willing to share digital identities. Finally, users with higher levels of trust had a more
positive disposition towards their willingness to share digital identities.
6.3.2 Contributions for SNS Users
This study contributes practical knowledge for social networking site users and
developers. SNS users have different approaches to sharing identity information,
with the results of this study demonstrating that users with high levels of privacy
concerns were less willing to share digital identities and vice-versa. Similarly, trust is
another important factor that impacts upon their likelihood of sharing information.
This study shows that users with high levels of trust were always willing to share
digital identities without concerns. At the same time, users‘ SNS experience and
knowledge were important factors in determining their willingness to share
information on SNS. This study suggests that SNS users‘ identities need to be
safeguarded in order to maintain their own privacy, and users need to choose
carefully the identities they want to share. It is recommended that users should read
the privacy policies and term and conditions of SNS before they join and share their
private information.
6.3.3 Contribution to SNS Developers
From the literature of this study it can be seen that millions of people are using SNS
daily—and for varying purposes. Different people have different experiences and
knowledge, which can be deciding factors that govern their willingness to share
identities on SNS. It can also be seen that higher privacy concerns of users have
115
negative effects upon their willingness to disclose personal information, however,
trust has a positive effect upon their willingness to share digital identities on SNS.
This statement suggests that SNS developers need to develop easy-to-understand
privacy policies and control settings for general users that increase the level of trust
that users feel about their participation in SNS.
6.4 Limitations of the Study and Future Research Opportunities
As in all research there are several limitations of this dissertation. The findings are
limited by certain choices and by the inevitable constraints imposed on the researcher
by circumstances during the time that this project was being conducted.
Nevertheless, some of the limitations discussed in this section have led the researcher
to perceive further opportunities for study in this area. A number of the limitations of
this study, as discussed in the sections below, point to research opportunities and
guidelines for future research.
First, there were limitations on data collection techniques used, sampling issues and
the time taken for data collection. Despite efforts to collect the data from different
sources, data could only be collected from one SNS, namely Facebook. The
consequences for this research of this outcome were that it limited the researcher‘s
ability to compare the experience of the respondents‘ use of Facebook with users of
other SNS. As such, an opportunity exists for further research to take the results of
this study and cross-compare them with results harvested from responses of users of
other SNS.
Second, there were concerns about the time taken for data collection. Because of
116
various constraints and the need to contact different owners and administrators of
different forums, it took two months to collect the data. The collection of the data
was dependent on the permission provided by the forum users and, because of the
voluntary nature of the survey, a great number of users ignored the links. This
limitation might be overcome if the survey was conducted by a different
organisation, such as a college or university. This could present a research
opportunity in the future to collect data because, from the research conducted for this
study, it became obvious that most of the SNS users were students, so there is an
opportunity for a more comprehensive data capture that might be organised under the
auspices of a university, college or similar institution.
This study conducted a survey with a mass population, and the sample used did not
include individuals in equal proportions based on demographical characteristics of
choice. This distribution may well increase the chance of bias in data collection and,
therefore, another recommendation for a further area of research might be the
distribution of a similar survey to a generalised population in a specific area for data
collection—which may yield different results. This may well provide a good
opportunity for further studies at a higher level.
Moreover, further research could focus on different aspects of trust that impact upon
SNS users‘ willingness to share their digital identities, for example, useful further
exploration could be made of the degree of sharing of information that takes place
between friends, as compared with the level of sharing between users whose only
connection to each other is via SNS.
Additionally, in general terms it is always mentioned that knowledge and experience
117
have an effect upon any decisions, whereas in this study it is shown that SNS
experience does not really impact on users‘ willingness for sharing digital identities.
In this scenario it will be preferable to collect data for qualitative data analysis by
conducting focus groups and expert interviews for further investigation to further
explore the findings.
6.5 Summary
The goal of this study was to develop a groundwork model of privacy and trust-based
users‘ willingness to share digital identities on SNS to explain the impact of these
factors on using SNS. For this purpose, this study developed a conceptual model
describing the privacy and trust based decision-making process and tested the
proposed model using a structural equation modelling technique on SNS user
behaviour data collected via a web survey. The result helps in understanding users‘
attitudes and the intentions of SNS in terms of the management of privacy concerns
and trust dimension; and assists in clarifying the implications for the development of
effective SNS services and application. The results of the measurement and structural
model tests lend support to the purposed research model. The structural model
provided a good fit to the data and most path coefficient in the model were found to
be statistically significant.
The major findings of this study are that if users have a high level of trust in SNS,
then they are much more willing to share digital identities on these sites. At the same
time, privacy is another major factor that impacts upon trust and willingness to share
digital identities. The results show that if users‘ privacy concerns have a negative
effect towards trust, they are less willing to share digital identities. Finally, this study
suggests that user privacy concerns are the major reason effecting users‘ willingness
118
to share digital identities.
This study has made a major contribution towards the literature by providing a better
understanding and new insights into privacy concerns; as well as trust and its effect
upon users‘ willingness to share their digital identities on SNS. This study also
contributes to knowledge of SNS users and developers regarding privacy concerns
and how this impacts upon sharing digital identities on SNS. Finally, this study
experienced some limitations in terms of methodology for collecting data and sample
size, where a forum was used to collect data, for which the results might vary when
compared to data collected from a more generalised population sample of SNS users.
There are great opportunities for further study by researchers via the collection of
data using different methodologies, such as collecting data from different sources
(e.g., universities, colleges and institutes). Also, there are other opportunities for
further study to develop current understanding of the direct effect of user knowledge
and experience upon privacy concerns and trust as it relates to users of SNS.
119
References
Ahn, GJ, Ko, M & Shehab, M 2008, 'Portable User-Centric Identity Management',
paper presented to The IFIP TC 11 23rd
International Information Security
Conference, Milano, Italy, September 7-10,2008.
Ahn, Y-Y, Han, S, Kwak, H, Moon, S & Jeong, H 2007, 'Analysis of topological
characteristics of huge online social networking services', paper presented to 16th
international conference on World Wide Web, Banff, Alberta, Canada, May 8–12,
2007.
Ajzen, I 1991, 'The theory of planned behavior', Organizational behavior and human
decision processes, vol. 50, no. 2, pp. 179-211.
Ajzen, I & Fishbein, M 1977, 'Attitude-behavior relations: A theoretical analysis and
review of empirical research', Psychological bulletin, vol. 84, no. 5, pp. 888 -918.
Ajzen, I & Fishbein, M 1980, Understanding attitudes and predicting social
behavior, illustrated edn, Prentice Hall, New York.
Albrechtslund, A 2008, 'Online social networking as participatory surveillance', First
Monday, vol. 13, no. 3, p. 3.
Altheide, DL 2000, 'Identity and the definition of the situation in a mass-mediated
context', Symbolic Interaction, vol. 23, no. 1, pp. 1-27.
Anderson, DR, Sweeney, DJ, Williams, TA, Freeman, J & Shoesmith, E 2010,
Statistics for business and economics, 2nd
edn, Cengage Learning, Hampshire, United
Kingdom.
Anderson, JC & Gerbing, DW 1988, 'Structural equation modeling in practice: A
review and recommended two-step approach', Psychological bulletin, vol. 103, no. 3,
pp. 411-23.
Andrews, D, Preece, J & Turoff, M 2001, 'A conceptual framework for demographic
groups resistant to online community interaction', paper presented to 34th
Annual
Hawaii International Conference on System Sciences, MD, USA January 3-6, 2001.
Andrews, DC 2002, 'Audience-specific online community design', Communications
of the ACM, vol. 45, no. 4, April, 2002, pp. 64-8.
Arbuckle, JL 2006, Amos 7.0 user's guide, 1ST
edn, SPSS Chicago, IL.
Bagozzi, RP & Edwards, JR 1998, 'A general approach for representing constructs in
organizational research', Organizational Research Methods, vol. 1, no. 1, pp. 45-87.
Barber, BM & Odean, T 2001, 'The internet and the investor', The Journal of
120
Economic Perspectives, vol. 15, no. 1, pp. 41-54.
Barnes, SB 2006, 'A privacy paradox: Social networking in the United States', First
Monday, vol. 11, no. 9, pp. 11-5.
Bateman, PJ, Pike, JC & Butler, BS 2011, 'To disclose or not: publicness in social
networking sites', Information Technology & People, vol. 24, no. 1, pp. 78-100.
Baumgartner, H & Homburg, C 1996, 'Applications of structural equation modeling
in marketing and consumer research: a review', International Journal of Research in
Marketing, vol. 13, no. 2, pp. 139-61.
Bentler, PM 1990, 'Comparative fit indexes in structural models', Psychological
bulletin, vol. 107, no. 2, pp. 238-46.
Bentler, PM 2007, 'On tests and indices for evaluating structural models',
Personality and Individual Differences, vol. 42, no. 5, pp. 825-9.
Bhattacherjee, A & Sanford, C 2006, 'Influence processes for information technology
acceptance: An elaboration likelihood model', MIS quarterly, vol. 30, no. 4, pp.
805-25.
Biesanz, JC & West, SG 2004, 'Towards Understanding Assessments of the Big
Five: Multitrait‐Multimethod Analyses of Convergent and Discriminant Validity
Across Measurement Occasion and Type of Observer', Journal of Personality, vol.
72, no. 4, pp. 845-76.
Binder, J, Howes, A & Sutcliffe, A 2009, 'The problem of conflicting social spheres:
effects of network structure on experienced tension in social network sites', paper
presented to 27th
international conference on Human factors in computing systems,
Boston, MA, USA, April 4-9, 2009.
Bonhard, P & Sasse, M 2006, '‘Knowing me, knowing you‘—Using profiles and
social networking to improve recommender systems', BT Technology Journal, vol.
24, no. 3, pp. 84-98.
Boyd, AW 2011, 'A Longitudinal Study of Social Media Privacy Behavior', PhD
thesis, Pace University,New York.
Boyd, DM & Ellison, NB 2008, 'Social network sites: Definition, history, and
scholarship', Journal of Computer Mediated Communication, vol. 13, no. 1, pp.
210-30.
Brass, DJ, Butterfield, KD & Skaggs, BC 1998, 'Relationships and unethical
behavior: A social network perspective', Academy of management review, vol. 23,
no. 1, pp. 14-31.
Bryman, A & Bell, E 2007, Business research methods, 2nd
edn, Oxford University
Press, New York.
121
Buote, VM, Wood, E & Pratt, M 2009, 'Exploring similarities and differences
between online and offline friendships: The role of attachment style', Computers in
Human Behavior, vol. 25, no. 2, pp. 560-7.
Byrne, BM 2001, 'Structural equation modeling with AMOS, EQS, and LISREL:
Comparative approaches to testing for the factorial validity of a measuring
instrument', International Journal of Testing, vol. 1, no. 1, pp. 55-86.
Campbell, DT & Fiske, DW 1959, 'Convergent and discriminant validation by the
multitrait-multimethod matrix', Psychological bulletin, vol. 56, no. 2, pp. 81-105.
Chang, HH & Chen, SW 2008, 'The impact of customer interface quality, satisfaction
and switching costs on e-loyalty: Internet experience as a moderator', Computers in
Human Behavior, vol. 24, no. 6, pp. 2927-44.
Chiu, PY, Cheung, CMK & Lee, MKO 2008, 'Online social networks: Why do ―we‖
use Facebook?', in JMC Miltiadis D. Lytras, Ernesto Damiani, Robert D. Tennyson,
David Avison, Gottfried Vossen, Patricia Ordonez De Pablos (ed.), The Open
Knowlege Society. A Computer Science and Information Systems Manifesto,
illustrated edn, Springer, Athens, Greece, vol. 19, pp. 67-74.
Christofides, E, Muise, A & Desmarais, S 2009, 'Information disclosure and control
on facebook: are they two sides of the same coin or two different processes?',
CyberPsychology & Behavior, vol. 12, no. 3, pp. 341-5.
Collis, J & Hussey, R 2009, Business Research-a Practical Guide for Undergraduate
and Postgraduate Students, 3rd
edn, Palgrave MacMillan Limited, London.
Constant, D, Sproull, L & Kiesler, S 1996, 'The kindness of strangers: The usefulness
of electronic weak ties for technical advice', Organization science, vol. 7, no. 2, pp.
119-35.
Cooper, DR & Schindler, PS 2003, Business research methods, 7th
edn, McGraw Hill
Internationa, New York.
Corritore, CL, Kracher, B & Wiedenbeck, S 2003, 'On-line trust: concepts, evolving
themes, a model', International Journal of Human-Computer Studies, vol. 58, no. 6,
pp. 737-58.
Crespo Herrero, A & Bosque, I 2008, 'The effect of innovativeness on the adoption
of B2C e-commerce: A model based on the Theory of Planned Behaviour',
Computers in Human Behavior, vol. 24, no. 6, pp. 2830-47.
Creswell, JW & Clark, VLP 2007, Designing and conducting mixed methods
research, 2nd
edn, Sage Publications, California, USA.
Creswell, JW, Hanson, WE, Clark, VLP & Morales, A 2007, 'Qualitative research
designs: Selection and implementation', Counseling Psychologist, vol. 35, no. 2,
p. 236.
122
Cronbach, LJ 1951, 'Coefficient alpha and the internal structure of tests',
Psychometrika, vol. 16, no. 3, pp. 297-334.
Crotty, M 1998, The foundations of social research: Meaning and perspective in the
research process, 1st edn, Sage Publications Ltd, California, USA.
Culnan, MJ & Bies, RJ 2003, 'Consumer privacy: Balancing economic and justice
considerations', Journal of Social Issues, vol. 59, no. 2, pp. 323-42.
Dahlen, M 2002, 'Learning the web: internet user experience and response to web
marketing in Sweden', Journal of Interactive Advertising, vol. 3, no. 1, pp. 25-33.
Davis, FD 1989, 'Perceived usefulness, perceived ease of use, and user acceptance of
information technology', MIS quarterly, vol. 13, no. 3, pp. 319-40.
Davis, FD 1993, 'User acceptance of information technology: system characteristics,
user perceptions and behavioral impacts', International Journal of Man Machine
Studes, vol. 38, no. 3, pp. 475-87.
Davison, RM, Clarke, R, Jeff, H, Langford, D & Kuo, B 2003, 'Information privacy
in a globally networked society: implications for IS research', Communications of the
Association for Information Systems, vol. 12, no. 1, pp. 341 - 65.
Debatin, B, Lovejoy, JP, Horn, AK & Hughes, BN 2009, 'Facebook and online
privacy: Attitudes, behaviors, and unintended consequences', Journal of Computer
Mediated Communication, vol. 15, no. 1, pp. 83-108.
DiMicco, J, Millen, DR, Geyer, W, Dugan, C, Brownholtz, B & Muller, M 2008,
'Motivations for social networking at work', paper presented to 2008 ACM
conference on Computer supported cooperative work, San Diego, CA, USA,
November 8–12, 2008.
Dinev, T & Hart, P 2006, 'An extended privacy calculus model for e-commerce
transactions', Information Systems Research, vol. 17, no. 1, p. 61.
Dwyer, C, Hiltz, SR & Passerini, K 2007, 'Trust and privacy concern within social
networking sites: A comparison of Facebook and MySpace', paper presented to 13th
Americas Conference on Information Systems, Keystone, Colorado, August 9-12
2007.
Dwyer, C, Hiltz, SR & Widmeyer, G 2008, 'Understanding development and usage
of social networking sites: The social software performance model', paper presented
to 41st Annual Hawaii International Conference on System Sciences, Waikoloa, HI,
January 7 -01- 2008.
Eagly, AH & Chaiken, S 1993, The psychology of attitudes, illustrated edn, Harcourt
Brace Jovanovich College Publishers, New York, USA.
Ellison, NB & Boyd, D 2007a, 'Social network sites: Definition, history, and
scholarship', Journal of Computer-Mediated Communication, vol. 13, no. 1,
123
pp. 210-30.
Ellison, NB, Steinfield, C & Lampe, C 2007b, 'The benefits of Facebook ―friends:‖
Social capital and college students‘ use of online social network sites', Journal of
Computer Mediated Communication, vol. 12, no. 4, pp. 1143-68.
Eriksen, A 2008, Glitch opens Bebo users' private details to others, The New
Zealand Herald, viewed 1st
August 2011, <http://www.nzherald.co.nz/nz/news
/article.cfm?c_id=1&objectid=10511780>.
Etezadi-Amoli, J & Farhoomand, AF 1996, 'A structural model of end user
computing satisfaction and user performance', Information & Management, vol. 30,
no. 2, pp. 65-73.
Facebook 2011, Press room, statistics., Palo Alto, CA, viewed March 23, 2011,
<http://www.facebook.com/press/info.php?statistics>.
Fischer, F 1998, 'Beyond empiricism: policy inquiry in post positivist perspective',
Policy Studies Journal, vol. 26, no. 1, pp. 129-46.
Fishbein, M & Ajzen, I 1975, Belief, attitude, intention and behavior -An
Introduction to Theory and Research, illustrated edn, Addison-Wesley series in
social psychology, Addison-Wesley, MA,USA.
Flicker, MH 2004, 'Securities trading on the Internet', in H Bidgoli (ed.), The Internet
Encyclopedia, 3rd
edn, Wiley Online Library, New Jersey, USA, vol. 3, p. 284.
Fogel, J & Nehmad, E 2009, 'Internet social network communities: Risk taking, trust,
and privacy concerns', Computers in Human Behavior, vol. 25, no. 1, pp. 153-60.
Forum, F 2011, What is Facebook Forum?, Palo Alto, CA, viewed 01 July 2011,
<http://www.thefacebookforum.net/>.
Fox, S & Purcell, K 2010, Chronic Disease and the Internet, Pew Research Center,
viewed 30th
November 2011, <http://pewinternet.org/Reports/2010/Chronic-
Disease.aspx>.
Fuchs, C 2010, 'studiVZ: social networking in the surveillance society', Ethics and
Information Technology, vol. 12, no. 2, pp. 171-85.
Fukuyama, F 1996, 'Trust Still Counts in a Virtual World', FORBES, vol. 158,
December 2, 1996, pp. 33-5.
Gangadharbatla, H 2010, 'Facebook me: Collective self-esteem, need to belong, and
internet self-efficacy as predictors of the igeneration's attitudes toward social
networking sites', Journal of Interactive Advertising, vol. 8, no. 2, pp. 5-15.
Ganley, D & Lampe, C 2009, 'The ties that bind: Social network principles in online
communities', Decision Support Systems, vol. 47, no. 3, pp. 266-74.
124
Gecas, V 1986, 'The motivational significance of self-concept for socialization
theory', in EJ Lawer (ed.), In Advances in group processes, JAI, Greenwich, vol. 3,
pp. 131-56.
Gefen, D & Straub, DW 2000, 'The relative importance of perceived ease of use in IS
adoption: A study of e-commerce adoption', Journal of the Association for
Information Systems, vol. 1, no. 1, pp. 1-30.
Gefen, D, Karahanna, E & Straub, DW 2003, 'Inexperience and experience with
online stores: The importance of TAM and trust', IEEE Transactions on Engineering
Management, vol. 50, no. 3, pp. 307-21.
George, JF 2002, 'Influences on the intent to make Internet purchases', Internet
Research, vol. 12, no. 2, pp. 165-80.
Gerbing, DW & Anderson, JC 1988, 'An updated paradigm for scale development
incorporating hnidimensionality and its assessment', Journal of Marketing Research,
vol. 25, no. 2, pp. 186-92.
Granovetter, M 1973, 'The strength of weak ties', American journal of Sociology, vol.
78, no. 6, pp. 1360-80.
Granovetter, M 1983, 'The strength of weak ties: A network theory revisited',
Sociological theory, vol. 1, no. 1, pp. 201-33.
Grayson, K & Ambler, T 1999, 'The dark side of long-term relationships in
marketing services', Journal of Marketing Research, vol. 36, no. 1, pp. 132-41.
Grewal, D, Munger, JL, Iyer, GR & Levy, M 2003, 'The influence of
internet‐retailing factors on price expectations', Psychology and Marketing, vol. 20,
no. 6, pp. 477-93.
Gross, R & Acquisti, A 2005, 'Information revelation and privacy in online social
networks', paper presented to 2005 ACM workshop on Privacy in the electronic
society, Alexandria, VA, USA, November 7, 2005.
Hair Jr, JF, Anderson, R, Tatham, R & Black, WC 1998, Multivariate Data Analysis,
5 th
edn, Prentice Hall, NJ.
Hair Jr, JF, Black, W, Babin, B, Anderson, R & Tatham, R 2006, Multivariate Data
Analysis, 6th
edn, Prentice Hall, NJ.
Hargittai, E 2008, 'Whose Space? Differences Among Users and Non Users of Social
Network Sites', Journal of Computer Mediated Communication, vol. 13, no. 1,
pp. 276-97.
Harrison, R & Thomas, M 2009, 'Identity in online communities: Social networking
sites and language learning', International Journal of Emerging Technologies &
Society, vol. 7, no. 2, pp. 109-24.
125
Hayduk, LA 1987, Structural equation modeling with LISREL: Essentials and
advances, 1st edn, The Johns Hopkins University Press, Baltimore, Maryland.
Hodge, MJ 2006, 'Fourth Amendment and Privacy Issues on the New Internet:
Facebook. com and Myspace. com, The', Southern Illinois University Law School
Journal, vol. 31, no. 3, pp. 95-122.
Hoffman, DL, Novak, TP & Peralta, M 1999, 'Building consumer trust online',
Communications of the ACM, vol. 42, no. 4, pp. 80-5.
Hoy, MG & Milne, G 2010, 'Gender Differences in Privacy-Related Measures for
Young Adult Facebook Users', Journal of Interactive Advertising, vol. 10, no. 2,
pp. 28-45.
Hu, Q & Ma, S 2010, 'Does Privacy Still Matter in the Era of Web 2.0? A Qualitative
Study of User Behavior towards Online Social Networking Activities', paper
presented to 14th
Pacific Asia Conference on Information Systems, Taipei, Taiwan,
July 9-12, 2010.
Joinson, AN, Reips, UD, Buchanan, T & Schofield, CBP 2010, 'Privacy, trust, and
self-disclosure online', Human–Computer Interaction, vol. 25, no. 1, pp. 1-24.
Joreskog, KG & Sorbom, D 1996, LISREL 8 user's reference guide, 2nd
edn,
Scientific Software International, Inc, Lincolnwood.
Kamel Boulos, MN & Wheeler, S 2007, 'The emerging Web 2.0 social software: an
enabling suite of sociable technologies in health and health care education1', Health
Information & Libraries Journal, vol. 24, no. 1, pp. 2-23.
Khosrow-Pour, M 2007, Web Technologies for Commerce and Services Online,
illustrated edn, IGI Publishing, Hershey, PA.
Kim, DJ, Steinfield, C & Lai, YJ 2008, 'Revisiting the role of web assurance seals in
business-to-consumer electronic commerce', Decision Support Systems, vol. 44, no.
4, pp. 1000-15.
Kline, RB 2010, Principles and practice of structural equation modeling, 3rd
edn,
The Guilford Press, NY.
Kripanont, N 2007, 'Examining a technology acceptance model of internet usage by
academics within Thai business schools', PhD thesis, Victoria University.
Lach, J 1999, 'The new gatekeepers', American Demographics, vol. 21, no. 6, pp.
41-2.
Lai, LSL & Turban, E 2008, 'Groups formation and operations in the Web 2.0
environment and social networks', Group Decision and Negotiation, vol. 17, no. 5,
pp. 387-402.
Leedy, PD & Ormrod, JE 2009, Practical research: Planning and design, 9th
126
edn, NJ.
Lenhart, A 2009, Adults and social network websites, Pew Internet and American
Life Project, viewed 29th
November 2011, <http://www.pewinternet.org/Reports
/2009/Adults-and-Social-Network-Websites.aspx>.
Lewis, K, Kaufman, J & Christakis, N 2008, 'The taste for privacy: An analysis of
college student privacy settings in an online social network', Journal of Computer
Mediated Communication, vol. 14, no. 1, pp. 79-100.
Liao, C, Chen, JL & Yen, DC 2007, 'Theory of planning behavior (TPB) and
customer satisfaction in the continued use of e-service: An integrated model',
Computers in Human Behavior, vol. 23, no. 6, pp. 2804-22.
Lin, HF 2006, 'Understanding behavioral intention to participate in virtual
communities', CyberPsychology & Behavior, vol. 9, no. 5, pp. 540-7.
LinkedIn 2011, About us, LinkedIn, viewed 28th
September 2011,
<http://press.linkedin.com/about>.
Liong, EL & Mejstad, T 2010, 'Privacy and Information Sharing on Social
Networking Sites', Master thesis, Lund University.
Liu, C, Marchewka, JT, Lu, J & Yu, CS 2005, 'Beyond concern--a privacy-trust-
behavioral intention model of electronic commerce', Information & Management,
vol. 42, no. 2, pp. 289-304.
Livingstone, S 2008, 'Taking risky opportunities in youthful content creation:
teenagers' use of social networking sites for intimacy, privacy and self-expression',
New Media & Society, vol. 10, no. 3, pp. 393-411.
Lo, J 2010, 'Privacy Concern, Locus of Control, and Salience in a Trust-Risk Model
of Information Disclosure on Social Networking Sites', paper presented to Americas
Conference on Information Systems, Lima, Peru, August 12-15, 2010.
Lo, J & Riemenschneider, C 2010, 'An Examination of Privacy Concerns and Trust
Entities in Determining Willingness to Disclose Personal Information on a Social
Networking Site', paper presented to Americas Conference on Information Systems,
Lima, Peru, August 12-15, 2010.
Luan, WS, Fung, NS, Nawawi, M & Hong, TS 2005, 'Experienced and
inexperienced Internet users among pre-service teachers: Their use and attitudes
toward the Internet', Educational Technology & Society, vol. 8, no. 1, pp. 90-103.
Luo, X 2002, 'Trust production and privacy concerns on the Internet: A framework
based on relationship marketing and social exchange theory', Industrial Marketing
Management, vol. 31, no. 2, pp. 111-8.
Madden, M & Fox, S 2006, Riding the waves of “Web 2.0.”, Pew Internet &
American Life Project, viewed 10th
November 2011,
127
<http://www.pewinternet.com/~/media/Files/Reports/2006/PIP_Web_2.0.pdf.>.
Malhotra, NK 2008, Marketing Research: An Applied Orientation, 5th
edn, Pearson
Education, India.
Malhotra, NK, Kim, SS & Agarwal, J 2004, 'Internet Users' Information Privacy
Concerns(IUIPC): The Construct, the Scale, and a Causal Model', Information
Systems Research, vol. 15, no. 4, pp. 336-55.
Manago, AM, Graham, MB, Greenfield, PM & Salimkhan, G 2008, 'Self-
presentation and gender on MySpace', Journal of Applied Developmental
Psychology, vol. 29, no. 6, pp. 446-58.
Mansfield-Devine, S 2008, 'Anti-social networking: exploiting the trusting
environment of Web 2.0', Network Security, vol. 2008, no. 11, pp. 4-7.
Maruyama, G 1998, Basics of structural equation modeling, 1st edn, Sage
Publications, Inc, Thousand Oaks, California
Mayer, RC, Davis, JH & Schoorman, FD 1995, 'An integrative model of
organizational trust', The Academy of Management Review, vol. 20, no. 3,
pp. 709-34.
Mayer, RC, Schoorman, FD & Davis, JH 2007, 'An integrative model of
organizational trust: Past, present, and future', The Academy of Management Review,
vol. 32, no. 2, pp. 344-54.
mc² 2011, my connected community, viewed 5th
November 2011,
<http://mc2.vicnet.net.au/>.
McDonald, RP & Ho, MHR 2002, 'Principles and practice in reporting structural
equation analyses', Psychological Methods, vol. 7, no. 1, p. 64.
McKnight, DH, Cummings, LL & Chervany, NL 1998, 'Initial trust formation in new
organizational relationships', The Academy of Management Review, vol. 23, no. 3,
pp. 473-90.
McKnight, DH, Choudhury, V & Kacmar, C 2003, 'Developing and validating trust
measures for e-commerce: An integrative typology', Information Systems Research,
vol. 13, no. 3, pp. 334-59.
Meinert, DB, Peterson, DK, Criswell, JR & Crossland, MD 2006a, 'Privacy policy
statements and consumer willingness to provide personal information', Journal of
Electronic Commerce in Organizations, vol. 4, no. 1, pp. 1-17.
Meinert, DB, Peterson, DK, Criswell, JR & Crossland, MD 2006b, 'Would
Regulation of Web Site Privacy Policy Statements Increase Consumer Trust?',
Informing Science Journal, vol. 9, no. 4, pp. 123-41.
Meredith, P 2006, Facebook and the Politics of Privacy, Mother Jones, viewed 5th
128
November 2011, <http://motherjones.com/politics/2006/09/facebook-and-politics-
privacy>.
Metzger, MJ 2004, 'Privacy, trust, and disclosure: Exploring barriers to electronic
commerce', Journal of Computer Mediated Communication, vol. 9, no. 4, pp. 00-.
Metzger, MJ & Docter, S 2003, 'Public opinion and policy initiatives for online
privacy protection', Journal of Broadcasting & Electronic Media, vol. 47, no. 3,
pp. 350-74.
Milne, GR & Culnan, MJ 2004, 'Strategies for reducing online privacy risks: Why
consumers read (or don't read) online privacy notices', Journal of Interactive
Marketing, vol. 18, no. 3, pp. 15-29.
Mislove, A, Marcon, M, Gummadi, KP, Druschel, P & Bhattacharjee, B 2007,
'Measurement and analysis of online social networks', paper presented to 7th
ACM
SIGCOMM conference on Internet measurement, San Diego, California, USA,
October 24-26, 2007.
Molok, NNA, Chang, S & Ahmad, A 2010, 'Information Leakage through Online
Social Networking: Opening the Doorway for Advanced Persistence Threats', paper
presented to 8th
Australian Information Security Management, Perth, Australia, 30th
November-2nd
December, 2010.
Mori, J, Sugiyama, T & Matsuo, Y 2005, 'Real-world oriented information sharing
using social networks', paper presented to GROUP 05 international ACM
SIGGROUP conference on Supporting group work, Sanibel Island, Florida, USA,
November 6-9, 2005.
MySpace 2011, Pressroom, fact sheet, Los Angeles, CA, viewed 23 March 2011,
<http://www.myspace.com/pressroom/2011/03/>.
Nahapiet, J & Ghoshal, S 1998, 'Social capital, intellectual capital, and the
organizational advantage', Academy of management review, vol. 23, no. 2, pp. 242-
66.
Nooteboom, B 2007, 'Social capital, institutions and trust', Review of Social
Economy, vol. 65, no. 1, pp. 29-53.
Novak, TP, Hoffman, DL & Yung, YF 2000, 'Measuring the customer experience in
online environments: A structural modeling approach', Marketing Science, vol. 19,
no. 1, pp. 22-42.
Nunnally, J 1994, Psychometric theory, 3rd
edn, Social Psychology, McGraw-Hill,
New York.
Nysveen, H & Pedersen, PE 2004, 'An exploratory study of customers' perception of
company web sites offering various interactive applications: moderating effects of
customers' internet experience', Decision Support Systems, vol. 37, no. 1, pp. 137-50.
129
Paine, C, Reips, UD, Stieger, S, Joinson, A & Buchanan, T 2007, 'Internet users'
perceptions of privacy concerns' and privacy actions'', International Journal of
Human-Computer Studies, vol. 65, no. 6, pp. 526-36.
Pallant, J 2011, SPSS Survival Manual, 4th
edn, Allen & Unwin, Crows Nest, NSW.
Park, JH, Konana, P, Gu, B & Man Leung, AC 2010, 'An investigation of
information sharing and seeking behaviors in virtual communities', paper presented
to International Conference on Information Systems, St. Louis, 8-01-2010.
Park, N, Kee, KF & Valenzuela, S 2009, 'Being immersed in social networking
environment: Facebook groups, uses and gratifications, and social outcomes',
CyberPsychology & Behavior, vol. 12, no. 6, pp. 729-33.
Paul, A, Hepworth, M, Kelly, B & Metcalfe, R 2007, 'What is Web 2. 0? Ideas,
technologies and implications for education ', Technology and Standards Watch, vol.
60, no. 1, pp. 2-64.
Pavlou, PA 2003, 'Consumer acceptance of electronic commerce: Integrating trust
and risk with the technology acceptance model', International Journal of Electronic
Commerce, vol. 7, no. 3, pp. 101-34.
Pavlou, PA & Fygenson, M 2006, 'Understanding and predicting electronic
commerce adoption: An extension of the theory of planned behavior', Management
Information Systems Quarterly, vol. 30, no. 1, pp. 115-43.
Peluchette, J & Karl, K 2008, 'Social networking profiles: An examination of student
attitudes regarding use and appropriateness of content', CyberPsychology &
Behavior, vol. 11, no. 1, pp. 95-7.
Peter, JP 1979, 'Reliability: A review of psychometric basics and recent marketing
practices', Journal of Marketing Research, vol. 16, no. 1, pp. 6-17.
Peter, JP 2001, 'Construct validity: a review of basic issues and marketing practices',
Journal of Marketing Research, vol. 18, no. 2, pp. 133-45.
Plant, R 2004, 'Online communities', Technology in Society, vol. 26, no. 1, pp. 51-65.
Ramgovind, S, Eloff, M & Smith, E 2010, 'The management of security in cloud
computing', paper presented to Information Security for South Africa (ISSA),
Sandton, Johannesburg August 2-4, 2010
Richards, DV 2006, 'Posting Personal Information on the Internet: A Case for
Changing the Legal Regime Created by Sec. 230 of the Communications Decency
Act', Texas Law Review, vol. 85, no. 5, pp. 1321-58.
Robards, B 2010, 'Randoms in my bedroom: Negotiating privacy and unsolicited
contact on social network sites', PRism, vol. 7, no. 3, p. 00.
Robinson, J, Shaverm, P & Wrightsman, L 1991, 'Criteria for scale selection and
130
evaluation ', in JP Robinson, PR Shaver & LS Wrightsman (eds), Measures of
personality and social psychological attitudes, Gulf Professional Publishing, San
Diego, CA, vol. 1, pp. 1-16.
Rosenblum, D 2007, 'What anyone can know: The privacy risks of social networking
sites', IEEE Security & Privacy, vol. 5, no. 3, pp. 40-9.
Rowe, M 2010, 'The credibility of digital identity information on the social web: a
user study', paper presented to WICOW‘10, Raleigh, North Carolina, USA, April 27,
2010.
Schumacker, RE & Lomax, RG 2004, A beginner's guide to structural equation
modeling, 2nd
edn, vol. 1, Lawrence Erlbaum, Mahwah, New Jersey.
Seale, C 1999, Quality in qualitative research, 1st edn, vol. 5, Qualitative inquiry,
SAGE publications, London.
Seddon, PB 1997, 'A respecification and extension of the DeLone and McLean
model of IS success', Information Systems Research, vol. 8, no. 3, pp. 240-53.
Sekaran, U 2003, Research methods for business: A skill building approach, 4th
edn,
CourseSmart, John Wiley and Sons, India.
Shafie, LA, Mansor, M, Osman, N, Nayan, S & Maesin, A 2011, 'Privacy, Trust and
Social Network Sites of University Students in Malaysia', Research Journal of
Internatıonal Studıes, no. 20, pp. 154 -62.
Sharma, S 1995, Applied multivariate techniques, 1st edn, John Wiley & Sons, New
York.
Shin, D-H 2010a, 'The effects of trust, security and privacy in social networking: A
security-based approach to understand the pattern of adoption', Interacting with
Computers, vol. 22, no. 5, pp. 428-38.
Shin, DH 2010b, 'Effect of Trust and Privacy Concerns on Social Networking: A
Trust-Based Acceptance Model for Social Networking Systems', paper presented to
International Communication Association, Suntec Singapore International
Convention & Exhibition Centre, Suntec City, Singapore, Jun 22, 2010
Son, JY & Kim, SS 2008, 'Internet users‘ information privacy-protective responses:
A taxonomy and a nomological model', Management Information Systems Quarterly,
vol. 32, no. 3, pp. 503-29.
Staksrud, E & Lobe, B 2010, Evaluation of the implementation of the safer social
networking principles for the EU, Europe's Information Society viewed 30th
August
2011,
<http://ec.europa.eu/information_society/activities/social_networking/eu_action/impl
ementation_princip/index_en.htm>.
Stone, GP 2005, 'Appearance and the self: A slightly revised version', in CE Dennis
131
Brissett (ed.), Life as theater: a dramaturgical sourcebook, 2nd
edn, Transaction
Publishers, New Jersey, vol. 2, pp. 141–62.
Sussman, SW & Siegal, WS 2003, 'Informational influence in organizations: An
integrated approach to knowledge adoption', Information Systems Research, vol. 14,
no. 1, pp. 47-65.
Taylor, S & Todd, P 1995a, 'Assessing IT usage: The role of prior experience', MIS
quarterly, vol. 19, no. 4, pp. 561-70.
Taylor, S & Todd, PA 1995b, 'Understanding information technology usage: A test
of competing models', Information Systems Research, vol. 6, no. 2, pp. 144-76.
Thelwall, M 2009, 'Social Network Sites:: Users and Uses', in M Zelkowitz (ed.),
Advances in computers, 1st edn, Academic Press, Massachusetts, vol. 76, pp. 19-73.
Thompson, B 2004, Exploratory and confirmatory factor analysis: Understanding
concepts and applications, 1at edn, American Psychological Association,
Washington, D.C
Thompson, B & Daniel, LG 1996, 'Factor analytic evidence for the construct validity
of score: a historical overview and some guidelines', Educational and Psychological
Measurement, vol. 56, no. 2, pp. 197-208.
Timm, DM & Duven, CJ 2008, 'Privacy and social networking sites', New Directions
for Student Services, vol. 2008, no. 124, pp. 89-101.
Trembath, L 2011, 'Determinants of success for online communities: An emprical
study', Doctor of Business Adminstration thesis, Swinburne University of
Technology, Melbourne.
Tufekci, Z 2008, 'Can you see me now? Audience and disclosure regulation in online
social network sites', Bulletin of Science, Technology & Society, vol. 28, no. 1, pp.
20-36.
Tull, DS & Hawkins, DI 1990, Marketing research: measurement and method: a text
with cases, 5th
edn, Macmillan, New York.
USC 2011, Digital Future Report, USC Annenberg School for Communication &
Journalism, viewed 15 th
August 2011,
<http://www.digitalcenter.org/pdf/2011_digital_future_final_release.pdf>.
Valenzuela, S, Park, N & Kee, KF 2009, 'Is There Social Capital in a Social Network
Site?: Facebook Use and College Students' Life Satisfaction, Trust, and
Participation1', Journal of Computer Mediated Communication, vol. 14, no. 4, pp.
875-901.
Van Slyke, C, Shim, J, Johnson, R & Jiang, JJ 2006, 'Concern for information
privacy and online consumer purchasing', Journal of the Association for Information
Systems, vol. 7, no. 6, pp. 415-44.
132
Veal, AJ 2005, Business research methods: A managerial approach, 2nd
edn, Pearson
Education, NSW.
Warren, SD & Brandeis, LD 1890, 'The right to privacy', Harvard law review, vol. 4,
no. 5, pp. 193-220.
Warshaw, PR & Davis, FD 1985, 'The accuracy of behavioral intention versus
behavioral expectation for predicting behavioral goals', The Journal of psychology,
vol. 119, no. 6, pp. 599-602.
Wasko, MML & Faraj, S 2005, 'Why should I share? Examining social capital and
knowledge contribution in electronic networks of practice', MIS quarterly, vol. 29,
no. 1, pp. 35-57.
Weik, P & Wahle, S 2010, 'Towards a Generic Identity Enabler for Telco Networks',
paper presented to 12 th
International Conference on Intelligence in Service Delivery
Networks (ICIN), Bordeaux, France, October 20-23 2008.
Wellman, B 1997, 'An electronic group is virtually a social network', in S Kiesler
(ed.), Culture of the Internet, 4th
edn, Routledge, New Jersey, pp. 179-205.
Wenger, E 2005, Communities of practice: Learning, meaning, and identity, 13th
edn, Cambridge University Press, New York.
Williams, S, Fleming, S, Lundqvist, K & Parslow, P 2010, 'Understanding your
digital identity', Learning Exchange, vol. 1, no. 1, pp. 1-5.
Windley, P 2005, Digital identity, 1st edn, O'Reilly Media, Inc., California.
Wixom, BH & Watson, HJ 2001, 'An empirical investigation of the factors affecting
data warehousing success', MIS quarterly, vol. 25, no. 1, pp. 17-41.
Wu, A, DiMicco, JM & Millen, DR 2010, 'Detecting professional versus personal
closeness using an enterprise social network site', paper presented to 28th
international conference on Human factors in computing systems, Atlanta, Georgia,
USA, April 10-15, 2010.
Xu, H 2009, 'consumer responses to the Introduction of Privacy Protection
Measures', International Journal of E-Business Research, vol. 5, no. 2, pp. 21-47.
Yahoo 2011, Connect with a world of people who share your passions, viewed 10th
November 2011, <http://au.groups.yahoo.com/>.
Ybarra, ML & Mitchell, KJ 2008, 'How risky are social networking sites? A
comparison of places online where youth sexual solicitation and harassment occurs',
American Academy of Pediatrics, vol. 121, no. 2, pp. 350-7.
Yu, T & Wu, G 2007, 'Determinants of Internet Shopping Behavior: An Application
of Reasoned Behaviour Theory', International Journal of Management, vol. 24, no.
133
4, p. 744.
Zarsky, TZ 2004, 'Desperately seeking solutions: Using implementation-based
solutions for the troubles of information privacy in the age of data mining and the
internet society', Maine Law Review, vol. 56, no. 1, pp. 13-59.
Zhang, Y & Tatipamula, M 2011, 'The freshman handbook: a hint for the server
placement of social networks', paper presented to 20th
international conference
companion on World wide web, Hyderabad, India, March 28–April 1, 2011.
Zhao, S 2005, 'The digital self: Through the looking glass of telecopresent others',
Symbolic Interaction, vol. 28, no. 3, pp. 387-405.
Zhao, S, Grasmuck, S & Martin, J 2008, 'Identity construction on Facebook: Digital
empowerment in anchored relationships', Computers in Human Behavior, vol. 24,
no. 5, pp. 1816-36.
Zikmund, WG, Griffin, M, Babin, BJ & Carr, JC 2009, Business research methods,
8th
edn, Cengage Learning, Mason, Ohio, USA.
134
Appendix A
Participant Information sheet
TO: Participants
TITLE OF PROJECT: Understanding impact of privacy concerns and trust in social networking
sites: Analysing user intentions towards willingness to share digital identities
RESEARCH TEAM: Sanjib Tiwari, MBSR Student, Faculty of Business and Law, University of
Southern Queensland, Phone: +62 7 46875775, email: [email protected]
Description
The purpose of this project is to address how user‘ privacy concerns and trust influence their
intentions towards willingness to share information in social networking sites.
The research team request your assistance because you are social networking sites user which eligible
as a participant of this project.
This project is being undertaken as part of a MBSR project for Sanjib Tiwari.
Participation
Your participation in this project is voluntary and non- participation will not affect you in any way.
You can withdraw from the project at any stage without comment or penalty. Your decision to
participate or not, or to withdraw from the project will not affect your current or future relationship
with the University of Southern Queensland.
This project involves the submission of anonymous (non-identifiable) material. Please note: it will not
be possible to withdraw your data once submitted.
It is expected your participation will take approximately 15 minutes of your time.
Please note: the data obtained from this project may be used at a later time for any research purpose.
Risks
There are no risks beyond day-to-day living associated with your participation in this project. This
survey is not expected to cause any discomfort or stress. If it does, you may discontinue taking the
survey. There is no compensation provided for taking this survey.
U n i v e r s i t y o f S o u t h e r n Q u e e n s l a n d
The University of Southern Queensland
Participant Information Sheet
135
Confidentiality
Any information obtained in connection with this project and that can identify you will remain
confidential. It will only be disclosed with your permission, subject to legal requirements. If you give
us your permission by signing the Consent Form, we plan to publish the results with my supervisor
to the Academic Journal.
In any publication, information will be provided in such a way that you cannot be identified.
All data received for this project will remain stored for a minimum of 5 years in secure facilities.
Consent to Participate
Please read this information sheet carefully so that you understand what the project involves. If you do
not understand any part of the project or require further information please contact the research teams
members named above. Please find the ethical acceptance letter from USQ ethics officer for the above
mention project.
The return of the completed anonymous survey is accepted as an indication of your
consent to participant in this project
Questions/further information about the project
You are encouraged to print this consent form and keep safe place you could contact the research
teams members named above if you have any questions or if you require further information about the
project. Now if you want to participate in this survey please click the below link or copy paste the link
in web browser:
http://usqbusiness.us.qualtrics.com/SE/?SID=SV_5vWRCqaLfUMG6Ve
Concerns/complaints regarding the conduct of the project
If you have any concerns or complaints about the ethical conduct of the project you may contact
the USQ Ethics Officer on +61 7 4631 2690 or email [email protected]. The Ethics Officer is
not connected with the project and can facilitate a resolution in an impartial manner.
Where the research may cause distress, independent 24 hour counselling services are available
through Lifeline on 13 11 14 from anywhere in Australia.
136
Appendix B
Ethics Approval
137
Appendix C
Information and informed consent statement
(online version)
Understanding the Social networking sites:
You are invited to participate in a research project that looks at‖ Understanding impact of privacy
concerns and trust in social networking sites: Analysing user intentions towards willingness to share
digital identities‖
You can participate in this project by completing the online questionnaire about the social networking
sites and by participating in this research you might find that you will learn something about impact of
privacy concerns and trust in SNS. It is expected that up to 200 members of a variety of social
networking sites will volunteer to participate in the study.
This questionnaire should take approximately 7 minutes to complete.
You will not be asked for any personal information that could identify you. This ensures your
anonymity, confidentially and privacy. To participate in this study you are required to be 18 years of
age, or older and be a member of at least one social networking sites.
Findings from the study, using aggregated data, will be reported in my thesis and possibly some co-
authored academic publications. The thesis will be submitted as partial fulfilment of a MBSR at
University of Southern Queensland.
Agreeing to complete this questionnaire is taken as your Informed Consent. Informed Consent means
you agree that your participation is voluntary and you understand that you are free to stop answering
the questions at any time. Only answers from completed questionnaires will be used in this study.
If you have any questions regarding this project please contact:
Sanjib Tiwari
MBSR Candidate
Dr. Jianming Yong
Principal Supervisor
Dr. Michael Lane
Associate Supervisor
This project has been approved by or on behalf of USQ Fast Track Human Research Ethics
Committee (FTHREC) in line with the National Statement on Ethical Conduct in Research Involving
Humans.
If you have any concerns or complaints about the ethical conduct of the project you may contact the
USQ Ethics Officer on +61 7 4631 2690 or email [email protected]. The Ethics Officer is not
connected with the project and can facilitate a resolution in an impartial manner.
Where the research may cause distress, independent 24 hour counselling services are available
through Lifeline on 13 11 14 from anywhere in Australia.
If you would like to assist us by completing the survey, please click on the start Link below:
http://usqbusiness.us.qualtrics.com/SE/?SID=SV_5vWRCqaLfUMG6Ve
138
Appendix D
Online Survey
A few glimpse of the online survey:
139
140
141
142
143
Appendix E
Statistical data analysis details
Reliability and Factor loading Statistics test using SPSS for Privacy Concerns
Reliability Statistics
Cronbach's Alpha
Cronbach's Alpha
Based on
Standardised
Items N of Items
.816 .817 5
Item-Total Statistics
Scale Mean if
Item Deleted
Scale Variance if
Item Deleted
Corrected Item-
Total Correlation
Squared Multiple
Correlation
Cronbach's Alpha
if Item Deleted
PC1 20.85 20.114 .564 .332 .793
PC2 20.54 20.237 .612 .387 .779
PC3 20.77 19.426 .652 .486 .767
PC4 20.71 18.961 .636 .464 .771
PC5 20.38 19.445 .573 .331 .791
Factor Matrixa
Factor
1
PC1 .611
PC2 .673
PC3 .770
PC4 .749
PC5 .626
Extraction Method: Maximum Likelihood.
a. 1 factors extracted. 3 iterations required.
144
Reliability and Factor loading Statistics test using SPSS for Trust
Reliability Statistics
Cronbach's Alpha
Cronbach's Alpha
Based on
Standardised
Items N of Items
.883 .886 5
Item-Total Statistics
Scale Mean if
Item Deleted
Scale Variance if
Item Deleted
Corrected Item-
Total Correlation
Squared Multiple
Correlation
Cronbach's Alpha
if Item Deleted
T1 18.25 23.086 .620 .406 .884
T2 18.12 22.051 .769 .603 .846
T3 18.23 22.267 .762 .627 .848
T4 18.37 22.896 .789 .652 .843
T6 17.80 24.953 .678 .505 .868
Factor Matrixa
Factor
1
T1 .660
T2 .810
T3 .840
T4 .857
T6 .736
Extraction Method: Maximum Likelihood.
a. 1 factors extracted. 4 iterations required.
145
Reliability and Factor loading Statistics test using SPSS for SNS Experience
Reliability Statistics
Cronbach's Alpha
Cronbach's Alpha
Based on
Standardised
Items N of Items
.707 ..709 4
Item-Total Statistics
Scale Mean if
Item Deleted
Scale Variance if
Item Deleted
Corrected Item-
Total Correlation
Squared Multiple
Correlation
Cronbach's Alpha
if Item Deleted
SNS1 18.77 2.413 .732 .021 .710
SNS2 19.38 1.939 ..690 .017 ..705
SNS3 10.52 3.108 .648 .042 ..701
SNS4 17.95 3.166 .685 .048 .707
Factor Matrixa
Factor
1
SNS1 .760
SNS2 .710
SNS3 .660
SNS4 .700
Extraction Method: Maximum Likelihood.
a. 1 factors extracted. 4 iterations required.
146
Reliability and Factor loading Statistics test using SPSS for Digital Identity
Reliability Statistics
Cronbach's Alpha
Cronbach's Alpha
Based on
Standardised
Items N of Items
.869 .879 6
Item-Total Statistics
Scale Mean if
Item Deleted
Scale Variance if
Item Deleted
Corrected Item-
Total Correlation
Squared Multiple
Correlation
Cronbach's Alpha
if Item Deleted
DI2 27.14 40.417 .468 .252 .892
DI7 26.36 41.609 .633 .472 .852
DI8 26.30 38.524 .777 .659 .827
DI9 26.27 38.498 .823 .721 .821
DI10 26.53 37.627 .737 .579 .833
DI13 25.79 43.169 .663 .477 .850
Factor Matrixa
Factor
1
DI2 .491
DI7 .701
DI8 .856
DI9 .909
DI10 .785
DI13 .711
Extraction Method: Maximum Likelihood.
a. 1 factors extracted. 4 iterations required.
147
Snapshot of AMOS first structural model with latent variable
The snapshot below describe that this is the model draw in AMOS 19 software tools.
Where we can see there are four variables Trust, Privacy, SNS and DI with number
of its latent variable
148
Snapshot of AMOS Unstandardised estimates structural model with latent variable
and factor loading
149
Snapshot of AMOS output of analysis summary for structural model
The snapshot below describe the analysis summary of the output of AMOS , where
we can see the time and model name which is analysed.
150
The snapshot below shows the summary of the variable used in model. The details of
variable are also shown in table below:
Variable counts (Group number 1)
Number of variables in your model: 46
Number of observed variables: 20
Number of unobserved variables: 26
Number of exogenous variables: 24
Number of endogenous variables: 22
151
Snapshot of AMOS output of structural model parameter summary
Weights Covariances Variances Means Intercepts Total
Fixed 26 0 0 0 0 26
Labeled 0 0 0 0 0 0
Unlabeled 20 1 24 0 0 45
Total 46 1 24 0 0 71
152
Snapshot of AMOS output of structural model
Number of distinct sample moments: 210
Number of distinct parameters to be estimated: 45
Degrees of freedom (210 - 45): 165
This Chi-square tests the null hypothesis that the over identified (reduced) model fits
the data as well as does a just-identified (full, saturated) model. In a just-identified
model there is a direct path (not through an intervening variable) from each variable
to each other variable. When we delete one or more of the paths we obtain an over
identified model. The non-significant Chi-square here indicated that the fit between
our over identified model and the data is not significantly worse than the fit between
the just-identified model and the data While one might argue that non-significance of
153
this Chi-square indicates that the reduced model fits the data well, even a well-fitting
reduced model will be significantly different from the full model if sample size is
sufficiently large. A good fitting model is one that can reproduce the original
variance-covariance matrix (or correlation matrix) from the path coefficients, in
much the same way that a good factor analytic solution can reproduce the original
correlation matrix with little error.
Estimate S.E. C.R. P Label
Trust <--- Privacy -.065 .106 -.619 .536 par_18
DI <--- SNS -1.124 1.101 -1.021 .307 par_1
DI <--- Trust .264 .089 2.962 .003 par_2
DI <--- Privacy -.270 .109 -2.474 .013 par_17
Estimate
154
Estimate
Trust <--- Privacy -.058
DI <--- SNS -.257
DI <--- Trust .282
DI <--- Privacy -.255
Estimate S.E. C.R. P Label
SNS <--> Privacy .025 .038 .667 .505 par_19
Below are the simple correlations between exogenous variables.
Estimate
SNS <--> Privacy .124
Top Related